166 files changed, 2348 insertions, 1221 deletions
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 6d3146c9e..f5ef0ef50 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -489,7 +489,7 @@ if (ENABLE_SDL2)
    if (YUZU_USE_BUNDLED_SDL2)
        # Detect toolchain and platform
        if ((MSVC_VERSION GREATER_EQUAL 1920 AND MSVC_VERSION LESS 1940) AND ARCHITECTURE_x86_64)
-            set(SDL2_VER "SDL2-2.0.18")
+            set(SDL2_VER "SDL2-2.28.0")
        else()
            message(FATAL_ERROR "No bundled SDL2 binaries for your toolchain. Disable YUZU_USE_BUNDLED_SDL2 and provide your own.")
        endif()
@@ -509,7 +509,7 @@ if (ENABLE_SDL2)
    elseif (YUZU_USE_EXTERNAL_SDL2)
        message(STATUS "Using SDL2 from externals.")
    else()
-        find_package(SDL2 2.0.18 REQUIRED)
+        find_package(SDL2 2.26.4 REQUIRED)
    endif()
 endif()
diff --git a/externals/SDL b/externals/SDL
-Subproject f17058b562c8a1090c0c996b42982721ace9090
+Subproject ffa78e6bead23e2ba3adf8ec2367ff2218d4343
diff --git a/src/android/app/src/main/java/org/yuzu/yuzu_emu/activities/EmulationActivity.kt b/src/android/app/src/main/java/org/yuzu/yuzu_emu/activities/EmulationActivity.kt
index 2b63388cc..ae665ed2e 100644
--- a/src/android/app/src/main/java/org/yuzu/yuzu_emu/activities/EmulationActivity.kt
+++ b/src/android/app/src/main/java/org/yuzu/yuzu_emu/activities/EmulationActivity.kt
@@ -27,13 +27,13 @@ import android.view.MotionEvent
 import android.view.Surface
 import android.view.View
 import android.view.inputmethod.InputMethodManager
+import android.widget.Toast
 import androidx.activity.viewModels
 import androidx.appcompat.app.AppCompatActivity
 import androidx.core.view.WindowCompat
 import androidx.core.view.WindowInsetsCompat
 import androidx.core.view.WindowInsetsControllerCompat
 import androidx.navigation.fragment.NavHostFragment
-import kotlin.math.roundToInt
 import org.yuzu.yuzu_emu.NativeLibrary
 import org.yuzu.yuzu_emu.R
 import org.yuzu.yuzu_emu.databinding.ActivityEmulationBinding
@@ -44,8 +44,10 @@ import org.yuzu.yuzu_emu.model.Game
 import org.yuzu.yuzu_emu.utils.ControllerMappingHelper
 import org.yuzu.yuzu_emu.utils.ForegroundService
 import org.yuzu.yuzu_emu.utils.InputHandler
+import org.yuzu.yuzu_emu.utils.MemoryUtil
 import org.yuzu.yuzu_emu.utils.NfcReader
 import org.yuzu.yuzu_emu.utils.ThemeHelper
+import kotlin.math.roundToInt
 class EmulationActivity : AppCompatActivity(), SensorEventListener {
    private lateinit var binding: ActivityEmulationBinding
@@ -104,6 +106,19 @@ class EmulationActivity : AppCompatActivity(), SensorEventListener {
        inputHandler = InputHandler()
        inputHandler.initialize()
+        val memoryUtil = MemoryUtil(this)
+        if (memoryUtil.isLessThan(8, MemoryUtil.Gb)) {
+            Toast.makeText(
+                this,
+                getString(
+                    R.string.device_memory_inadequate,
+                    memoryUtil.getDeviceRAM(),
+                    "8 ${getString(R.string.memory_gigabyte)}"
+                ),
+                Toast.LENGTH_LONG
+            ).show()
+        }
        // Start a foreground service to prevent the app from getting killed in the background
        val startIntent = Intent(this, ForegroundService::class.java)
        startForegroundService(startIntent)
diff --git a/src/android/app/src/main/java/org/yuzu/yuzu_emu/fragments/LongMessageDialogFragment.kt b/src/android/app/src/main/java/org/yuzu/yuzu_emu/fragments/LongMessageDialogFragment.kt
new file mode 100644
index 000000000..b29b627e9
--- /dev/null
+++ b/src/android/app/src/main/java/org/yuzu/yuzu_emu/fragments/LongMessageDialogFragment.kt
@@ -0,0 +1,62 @@
+// SPDX-FileCopyrightText: 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+package org.yuzu.yuzu_emu.fragments
+import android.app.Dialog
+import android.content.Intent
+import android.net.Uri
+import android.os.Bundle
+import androidx.fragment.app.DialogFragment
+import com.google.android.material.dialog.MaterialAlertDialogBuilder
+import org.yuzu.yuzu_emu.R
+class LongMessageDialogFragment : DialogFragment() {
+    override fun onCreateDialog(savedInstanceState: Bundle?): Dialog {
+        val titleId = requireArguments().getInt(TITLE)
+        val description = requireArguments().getString(DESCRIPTION)
+        val helpLinkId = requireArguments().getInt(HELP_LINK)
+        val dialog = MaterialAlertDialogBuilder(requireContext())
+            .setPositiveButton(R.string.close, null)
+            .setTitle(titleId)
+            .setMessage(description)
+        if (helpLinkId != 0) {
+            dialog.setNeutralButton(R.string.learn_more) { _, _ ->
+                openLink(getString(helpLinkId))
+            }
+        }
+        return dialog.show()
+    }
+    private fun openLink(link: String) {
+        val intent = Intent(Intent.ACTION_VIEW, Uri.parse(link))
+        startActivity(intent)
+    }
+    companion object {
+        const val TAG = "LongMessageDialogFragment"
+        private const val TITLE = "Title"
+        private const val DESCRIPTION = "Description"
+        private const val HELP_LINK = "Link"
+        fun newInstance(
+            titleId: Int,
+            description: String,
+            helpLinkId: Int = 0
+        ): LongMessageDialogFragment {
+            val dialog = LongMessageDialogFragment()
+            val bundle = Bundle()
+            bundle.apply {
+                putInt(TITLE, titleId)
+                putString(DESCRIPTION, description)
+                putInt(HELP_LINK, helpLinkId)
+            }
+            dialog.arguments = bundle
+            return dialog
+        }
+    }
+}
diff --git a/src/android/app/src/main/java/org/yuzu/yuzu_emu/ui/main/MainActivity.kt b/src/android/app/src/main/java/org/yuzu/yuzu_emu/ui/main/MainActivity.kt
index cc1d87f1b..3086cfad3 100644
--- a/src/android/app/src/main/java/org/yuzu/yuzu_emu/ui/main/MainActivity.kt
+++ b/src/android/app/src/main/java/org/yuzu/yuzu_emu/ui/main/MainActivity.kt
@@ -4,6 +4,7 @@
 package org.yuzu.yuzu_emu.ui.main
 import android.content.Intent
+import android.net.Uri
 import android.os.Bundle
 import android.view.View
 import android.view.ViewGroup.MarginLayoutParams
@@ -42,6 +43,7 @@ import org.yuzu.yuzu_emu.features.settings.model.SettingsViewModel
 import org.yuzu.yuzu_emu.features.settings.ui.SettingsActivity
 import org.yuzu.yuzu_emu.features.settings.utils.SettingsFile
 import org.yuzu.yuzu_emu.fragments.IndeterminateProgressDialogFragment
+import org.yuzu.yuzu_emu.fragments.LongMessageDialogFragment
 import org.yuzu.yuzu_emu.fragments.MessageDialogFragment
 import org.yuzu.yuzu_emu.model.GamesViewModel
 import org.yuzu.yuzu_emu.model.HomeViewModel
@@ -481,62 +483,110 @@ class MainActivity : AppCompatActivity(), ThemeProvider {
            }
        }
-    val installGameUpdate =
+    val installGameUpdate = registerForActivityResult(
-        registerForActivityResult(ActivityResultContracts.OpenDocument()) {
+        ActivityResultContracts.OpenMultipleDocuments()
-            if (it == null) {
+    ) { documents: List<Uri> ->
-                return@registerForActivityResult
+        if (documents.isNotEmpty()) {
-            }
            IndeterminateProgressDialogFragment.newInstance(
                this@MainActivity,
                R.string.install_game_content
            ) {
-                val result = NativeLibrary.installFileToNand(it.toString())
+                var installSuccess = 0
+                var installOverwrite = 0
+                var errorBaseGame = 0
+                var errorExtension = 0
+                var errorOther = 0
+                var errorTotal = 0
                lifecycleScope.launch {
-                    withContext(Dispatchers.Main) {
+                    documents.forEach {
-                        when (result) {
+                        when (NativeLibrary.installFileToNand(it.toString())) {
                            NativeLibrary.InstallFileToNandResult.Success -> {
-                                Toast.makeText(
+                                installSuccess += 1
-                                    applicationContext,
-                                    R.string.install_game_content_success,
-                                    Toast.LENGTH_SHORT
-                                ).show()
                            }
                            NativeLibrary.InstallFileToNandResult.SuccessFileOverwritten -> {
-                                Toast.makeText(
+                                installOverwrite += 1
-                                    applicationContext,
-                                    R.string.install_game_content_success_overwrite,
-                                    Toast.LENGTH_SHORT
-                                ).show()
                            }
                            NativeLibrary.InstallFileToNandResult.ErrorBaseGame -> {
-                                MessageDialogFragment.newInstance(
+                                errorBaseGame += 1
-                                    R.string.install_game_content_failure,
-                                    R.string.install_game_content_failure_base
-                                ).show(supportFragmentManager, MessageDialogFragment.TAG)
                            }
                            NativeLibrary.InstallFileToNandResult.ErrorFilenameExtension -> {
-                                MessageDialogFragment.newInstance(
+                                errorExtension += 1
-                                    R.string.install_game_content_failure,
-                                    R.string.install_game_content_failure_file_extension,
-                                    R.string.install_game_content_help_link
-                                ).show(supportFragmentManager, MessageDialogFragment.TAG)
                            }
                            else -> {
-                                MessageDialogFragment.newInstance(
+                                errorOther += 1
-                                    R.string.install_game_content_failure,
-                                    R.string.install_game_content_failure_description,
-                                    R.string.install_game_content_help_link
-                                ).show(supportFragmentManager, MessageDialogFragment.TAG)
                            }
                        }
                    }
+                    withContext(Dispatchers.Main) {
+                        val separator = System.getProperty("line.separator") ?: "\n"
+                        val installResult = StringBuilder()
+                        if (installSuccess > 0) {
+                            installResult.append(
+                                getString(
+                                    R.string.install_game_content_success_install,
+                                    installSuccess
+                                )
+                            )
+                            installResult.append(separator)
+                        }
+                        if (installOverwrite > 0) {
+                            installResult.append(
+                                getString(
+                                    R.string.install_game_content_success_overwrite,
+                                    installOverwrite
+                                )
+                            )
+                            installResult.append(separator)
+                        }
+                        errorTotal = errorBaseGame + errorExtension + errorOther
+                        if (errorTotal > 0) {
+                            installResult.append(separator)
+                            installResult.append(
+                                getString(
+                                    R.string.install_game_content_failed_count,
+                                    errorTotal
+                                )
+                            )
+                            installResult.append(separator)
+                            if (errorBaseGame > 0) {
+                                installResult.append(separator)
+                                installResult.append(
+                                    getString(R.string.install_game_content_failure_base)
+                                )
+                                installResult.append(separator)
+                            }
+                            if (errorExtension > 0) {
+                                installResult.append(separator)
+                                installResult.append(
+                                    getString(R.string.install_game_content_failure_file_extension)
+                                )
+                                installResult.append(separator)
+                            }
+                            if (errorOther > 0) {
+                                installResult.append(
+                                    getString(R.string.install_game_content_failure_description)
+                                )
+                                installResult.append(separator)
+                            }
+                            LongMessageDialogFragment.newInstance(
+                                R.string.install_game_content_failure,
+                                installResult.toString().trim(),
+                                R.string.install_game_content_help_link
+                            ).show(supportFragmentManager, LongMessageDialogFragment.TAG)
+                        } else {
+                            LongMessageDialogFragment.newInstance(
+                                R.string.install_game_content_success,
+                                installResult.toString().trim()
+                            ).show(supportFragmentManager, LongMessageDialogFragment.TAG)
+                        }
+                    }
                }
-                return@newInstance result
+                return@newInstance installSuccess + installOverwrite + errorTotal
            }.show(supportFragmentManager, IndeterminateProgressDialogFragment.TAG)
        }
+    }
 }
diff --git a/src/android/app/src/main/java/org/yuzu/yuzu_emu/utils/MemoryUtil.kt b/src/android/app/src/main/java/org/yuzu/yuzu_emu/utils/MemoryUtil.kt
new file mode 100644
index 000000000..18e5fa0b0
--- /dev/null
+++ b/src/android/app/src/main/java/org/yuzu/yuzu_emu/utils/MemoryUtil.kt
@@ -0,0 +1,59 @@
+// SPDX-FileCopyrightText: 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+package org.yuzu.yuzu_emu.utils
+import android.app.ActivityManager
+import android.content.Context
+import org.yuzu.yuzu_emu.R
+import java.util.Locale
+class MemoryUtil(val context: Context) {
+    private val Long.floatForm: String
+        get() = String.format(Locale.ROOT, "%.2f", this.toDouble())
+    private fun bytesToSizeUnit(size: Long): String {
+        return when {
+            size < Kb -> "${size.floatForm} ${context.getString(R.string.memory_byte)}"
+            size < Mb -> "${(size / Kb).floatForm} ${context.getString(R.string.memory_kilobyte)}"
+            size < Gb -> "${(size / Mb).floatForm} ${context.getString(R.string.memory_megabyte)}"
+            size < Tb -> "${(size / Gb).floatForm} ${context.getString(R.string.memory_gigabyte)}"
+            size < Pb -> "${(size / Tb).floatForm} ${context.getString(R.string.memory_terabyte)}"
+            size < Eb -> "${(size / Pb).floatForm} ${context.getString(R.string.memory_petabyte)}"
+            else -> "${(size / Eb).floatForm} ${context.getString(R.string.memory_exabyte)}"
+        }
+    }
+    private val totalMemory =
+        with(context.getSystemService(Context.ACTIVITY_SERVICE) as ActivityManager) {
+            val memInfo = ActivityManager.MemoryInfo()
+            getMemoryInfo(memInfo)
+            memInfo.totalMem
+        }
+    fun isLessThan(minimum: Int, size: Long): Boolean {
+        return when (size) {
+            Kb -> totalMemory < Mb && totalMemory < minimum
+            Mb -> totalMemory < Gb && (totalMemory / Mb) < minimum
+            Gb -> totalMemory < Tb && (totalMemory / Gb) < minimum
+            Tb -> totalMemory < Pb && (totalMemory / Tb) < minimum
+            Pb -> totalMemory < Eb && (totalMemory / Pb) < minimum
+            Eb -> totalMemory / Eb < minimum
+            else -> totalMemory < Kb && totalMemory < minimum
+        }
+    }
+    fun getDeviceRAM(): String {
+        return bytesToSizeUnit(totalMemory)
+    }
+    companion object {
+        const val Kb: Long = 1024
+        const val Mb = Kb * 1024
+        const val Gb = Mb * 1024
+        const val Tb = Gb * 1024
+        const val Pb = Tb * 1024
+        const val Eb = Pb * 1024
+    }
+}
diff --git a/src/android/app/src/main/jni/native.cpp b/src/android/app/src/main/jni/native.cpp
index 6688416d6..d576aac50 100644
--- a/src/android/app/src/main/jni/native.cpp
+++ b/src/android/app/src/main/jni/native.cpp
@@ -238,6 +238,7 @@ public:
        m_software_keyboard = android_keyboard.get();
        m_system.SetShuttingDown(false);
        m_system.ApplySettings();
+        Settings::LogSettings();
        m_system.HIDCore().ReloadInputDevices();
        m_system.SetAppletFrontendSet({
            nullptr,                     // Amiibo Settings
diff --git a/src/android/app/src/main/res/values/strings.xml b/src/android/app/src/main/res/values/strings.xml
index 381dfbc3b..af7450619 100644
--- a/src/android/app/src/main/res/values/strings.xml
+++ b/src/android/app/src/main/res/values/strings.xml
@@ -104,12 +104,14 @@
    <string name="share_log_missing">No log file found</string>
    <string name="install_game_content">Install game content</string>
    <string name="install_game_content_description">Install game updates or DLC</string>
-    <string name="install_game_content_failure">Error installing file to NAND</string>
+    <string name="install_game_content_failure">Error installing file(s) to NAND</string>
-    <string name="install_game_content_failure_description">Game content installation failed. Please ensure content is valid and that the prod.keys file is installed.</string>
+    <string name="install_game_content_failure_description">Please ensure content(s) are valid and that the prod.keys file is installed.</string>
-    <string name="install_game_content_failure_base">Installation of base games isn\'t permitted in order to avoid possible conflicts. Please select an update or DLC instead.</string>
+    <string name="install_game_content_failure_base">Installation of base games isn\'t permitted in order to avoid possible conflicts.</string>
-    <string name="install_game_content_failure_file_extension">The selected file type is not supported. Only NSP and XCI content is supported for this action. Please verify the game content is valid.</string>
+    <string name="install_game_content_failure_file_extension">Only NSP and XCI content is supported. Please verify the game content(s) are valid.</string>
-    <string name="install_game_content_success">Game content installed successfully</string>
+    <string name="install_game_content_failed_count">%1$d installation error(s)</string>
-    <string name="install_game_content_success_overwrite">Game content was overwritten successfully</string>
+    <string name="install_game_content_success">Game content(s) installed successfully</string>
+    <string name="install_game_content_success_install">%1$d installed successfully</string>
+    <string name="install_game_content_success_overwrite">%1$d overwritten successfully</string>
    <string name="install_game_content_help_link">https://yuzu-emu.org/help/quickstart/#dumping-installed-updates</string>
    <!-- About screen strings -->
@@ -270,6 +272,7 @@
    <string name="fatal_error">Fatal Error</string>
    <string name="fatal_error_message">A fatal error occurred. Check the log for details.\nContinuing emulation may result in crashes and bugs.</string>
    <string name="performance_warning">Turning off this setting will significantly reduce emulation performance! For the best experience, it is recommended that you leave this setting enabled.</string>
+    <string name="device_memory_inadequate">Device RAM: %1$s\nRecommended: %2$s</string>
    <!-- Region Names -->
    <string name="region_japan">Japan</string>
@@ -300,6 +303,15 @@
    <string name="language_traditional_chinese">Traditional Chinese (正體中文)</string>
    <string name="language_brazilian_portuguese">Brazilian Portuguese (Português do Brasil)</string>
+    <!-- Memory Sizes -->
+    <string name="memory_byte">Byte</string>
+    <string name="memory_kilobyte">KB</string>
+    <string name="memory_megabyte">MB</string>
+    <string name="memory_gigabyte">GB</string>
+    <string name="memory_terabyte">TB</string>
+    <string name="memory_petabyte">PB</string>
+    <string name="memory_exabyte">EB</string>
    <!-- Renderer APIs -->
    <string name="renderer_vulkan">Vulkan</string>
    <string name="renderer_none">None</string>
diff --git a/src/audio_core/device/audio_buffers.h b/src/audio_core/device/audio_buffers.h
index 15082f6c6..5d8ed0ef7 100644
--- a/src/audio_core/device/audio_buffers.h
+++ b/src/audio_core/device/audio_buffers.h
@@ -7,6 +7,7 @@
 #include <mutex>
 #include <span>
 #include <vector>
+#include <boost/container/static_vector.hpp>
 #include "audio_buffer.h"
 #include "audio_core/device/device_session.h"
@@ -48,7 +49,7 @@ public:
     *
     * @param out_buffers - The buffers which were registered.
     */
-    void RegisterBuffers(std::vector<AudioBuffer>& out_buffers) {
+    void RegisterBuffers(boost::container::static_vector<AudioBuffer, N>& out_buffers) {
        std::scoped_lock l{lock};
        const s32 to_register{std::min(std::min(appended_count, BufferAppendLimit),
                                       BufferAppendLimit - registered_count)};
@@ -162,7 +163,8 @@ public:
     * @param max_buffers     - Maximum number of buffers to released.
     * @return The number of buffers released.
     */
-    u32 GetRegisteredAppendedBuffers(std::vector<AudioBuffer>& buffers_flushed, u32 max_buffers) {
+    u32 GetRegisteredAppendedBuffers(
+        boost::container::static_vector<AudioBuffer, N>& buffers_flushed, u32 max_buffers) {
        std::scoped_lock l{lock};
        if (registered_count + appended_count == 0) {
            return 0;
@@ -270,7 +272,7 @@ public:
     */
    bool FlushBuffers(u32& buffers_released) {
        std::scoped_lock l{lock};
-        std::vector<AudioBuffer> buffers_flushed{};
+        boost::container::static_vector<AudioBuffer, N> buffers_flushed{};
        buffers_released = GetRegisteredAppendedBuffers(buffers_flushed, append_limit);
diff --git a/src/audio_core/device/device_session.cpp b/src/audio_core/device/device_session.cpp
index b5c0ef0e6..86811fcb8 100644
--- a/src/audio_core/device/device_session.cpp
+++ b/src/audio_core/device/device_session.cpp
@@ -79,7 +79,7 @@ void DeviceSession::ClearBuffers() {
    }
 }
-void DeviceSession::AppendBuffers(std::span<const AudioBuffer> buffers) const {
+void DeviceSession::AppendBuffers(std::span<const AudioBuffer> buffers) {
    for (const auto& buffer : buffers) {
        Sink::SinkBuffer new_buffer{
            .frames = buffer.size / (channel_count * sizeof(s16)),
@@ -88,13 +88,13 @@ void DeviceSession::AppendBuffers(std::span<const AudioBuffer> buffers) const {
            .consumed = false,
        };
+        tmp_samples.resize_destructive(buffer.size / sizeof(s16));
        if (type == Sink::StreamType::In) {
-            std::vector<s16> samples{};
+            stream->AppendBuffer(new_buffer, tmp_samples);
-            stream->AppendBuffer(new_buffer, samples);
        } else {
-            std::vector<s16> samples(buffer.size / sizeof(s16));
+            system.ApplicationMemory().ReadBlockUnsafe(buffer.samples, tmp_samples.data(),
-            system.ApplicationMemory().ReadBlockUnsafe(buffer.samples, samples.data(), buffer.size);
+                                                       buffer.size);
-            stream->AppendBuffer(new_buffer, samples);
+            stream->AppendBuffer(new_buffer, tmp_samples);
        }
    }
 }
diff --git a/src/audio_core/device/device_session.h b/src/audio_core/device/device_session.h
index 75f766c68..7d52f362d 100644
--- a/src/audio_core/device/device_session.h
+++ b/src/audio_core/device/device_session.h
@@ -10,6 +10,7 @@
 #include "audio_core/common/common.h"
 #include "audio_core/sink/sink.h"
+#include "common/scratch_buffer.h"
 #include "core/hle/service/audio/errors.h"
 namespace Core {
@@ -62,7 +63,7 @@ public:
     *
     * @param buffers - The buffers to play.
     */
-    void AppendBuffers(std::span<const AudioBuffer> buffers) const;
+    void AppendBuffers(std::span<const AudioBuffer> buffers);
    /**
     * (Audio In only) Pop samples from the backend, and write them back to this buffer's address.
@@ -146,8 +147,8 @@ private:
    std::shared_ptr<Core::Timing::EventType> thread_event;
    /// Is this session initialised?
    bool initialized{};
-    /// Buffer queue
+    /// Temporary sample buffer
-    std::vector<AudioBuffer> buffer_queue{};
+    Common::ScratchBuffer<s16> tmp_samples{};
 };
 } // namespace AudioCore
diff --git a/src/audio_core/in/audio_in_system.cpp b/src/audio_core/in/audio_in_system.cpp
index e23e51758..579129121 100644
--- a/src/audio_core/in/audio_in_system.cpp
+++ b/src/audio_core/in/audio_in_system.cpp
@@ -2,6 +2,7 @@
 // SPDX-License-Identifier: GPL-2.0-or-later
 #include <mutex>
 #include "audio_core/audio_event.h"
 #include "audio_core/audio_manager.h"
 #include "audio_core/in/audio_in_system.h"
@@ -89,7 +90,7 @@ Result System::Start() {
    session->Start();
    state = State::Started;
-    std::vector<AudioBuffer> buffers_to_flush{};
+    boost::container::static_vector<AudioBuffer, BufferCount> buffers_to_flush{};
    buffers.RegisterBuffers(buffers_to_flush);
    session->AppendBuffers(buffers_to_flush);
    session->SetRingSize(static_cast<u32>(buffers_to_flush.size()));
@@ -134,7 +135,7 @@ bool System::AppendBuffer(const AudioInBuffer& buffer, const u64 tag) {
 void System::RegisterBuffers() {
    if (state == State::Started) {
-        std::vector<AudioBuffer> registered_buffers{};
+        boost::container::static_vector<AudioBuffer, BufferCount> registered_buffers{};
        buffers.RegisterBuffers(registered_buffers);
        session->AppendBuffers(registered_buffers);
    }
diff --git a/src/audio_core/out/audio_out_system.cpp b/src/audio_core/out/audio_out_system.cpp
index bd13f7219..0adf64bd3 100644
--- a/src/audio_core/out/audio_out_system.cpp
+++ b/src/audio_core/out/audio_out_system.cpp
@@ -89,7 +89,7 @@ Result System::Start() {
    session->Start();
    state = State::Started;
-    std::vector<AudioBuffer> buffers_to_flush{};
+    boost::container::static_vector<AudioBuffer, BufferCount> buffers_to_flush{};
    buffers.RegisterBuffers(buffers_to_flush);
    session->AppendBuffers(buffers_to_flush);
    session->SetRingSize(static_cast<u32>(buffers_to_flush.size()));
@@ -134,7 +134,7 @@ bool System::AppendBuffer(const AudioOutBuffer& buffer, u64 tag) {
 void System::RegisterBuffers() {
    if (state == State::Started) {
-        std::vector<AudioBuffer> registered_buffers{};
+        boost::container::static_vector<AudioBuffer, BufferCount> registered_buffers{};
        buffers.RegisterBuffers(registered_buffers);
        session->AppendBuffers(registered_buffers);
    }
diff --git a/src/audio_core/renderer/adsp/adsp.cpp b/src/audio_core/renderer/adsp/adsp.cpp
index 74772fc50..b1db31e93 100644
--- a/src/audio_core/renderer/adsp/adsp.cpp
+++ b/src/audio_core/renderer/adsp/adsp.cpp
@@ -7,7 +7,6 @@
 #include "common/logging/log.h"
 #include "core/core.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 #include "core/memory.h"
 namespace AudioCore::AudioRenderer::ADSP {
diff --git a/src/audio_core/renderer/adsp/audio_renderer.cpp b/src/audio_core/renderer/adsp/audio_renderer.cpp
index 8bc39f9f9..9ca716b60 100644
--- a/src/audio_core/renderer/adsp/audio_renderer.cpp
+++ b/src/audio_core/renderer/adsp/audio_renderer.cpp
@@ -13,7 +13,6 @@
 #include "common/thread.h"
 #include "core/core.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 MICROPROFILE_DEFINE(Audio_Renderer, "Audio", "DSP", MP_RGB(60, 19, 97));
@@ -144,6 +143,7 @@ void AudioRenderer::ThreadFunc(std::stop_token stop_token) {
    mailbox->ADSPSendMessage(RenderMessage::AudioRenderer_InitializeOK);
+    // 0.12 seconds (2304000 / 19200000)
    constexpr u64 max_process_time{2'304'000ULL};
    while (!stop_token.stop_requested()) {
@@ -184,8 +184,7 @@ void AudioRenderer::ThreadFunc(std::stop_token stop_token) {
                    u64 max_time{max_process_time};
                    if (index == 1 && command_buffer.applet_resource_user_id ==
                                          mailbox->GetCommandBuffer(0).applet_resource_user_id) {
-                        max_time = max_process_time -
+                        max_time = max_process_time - render_times_taken[0];
-                                   Core::Timing::CyclesToNs(render_times_taken[0]).count();
                        if (render_times_taken[0] > max_process_time) {
                            max_time = 0;
                        }
diff --git a/src/audio_core/renderer/adsp/command_list_processor.cpp b/src/audio_core/renderer/adsp/command_list_processor.cpp
index 7a300d216..3a0f1ae38 100644
--- a/src/audio_core/renderer/adsp/command_list_processor.cpp
+++ b/src/audio_core/renderer/adsp/command_list_processor.cpp
@@ -9,7 +9,6 @@
 #include "common/settings.h"
 #include "core/core.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 #include "core/memory.h"
 namespace AudioCore::AudioRenderer::ADSP {
diff --git a/src/audio_core/renderer/command/data_source/decode.cpp b/src/audio_core/renderer/command/data_source/decode.cpp
index ff5d31bd6..f45933203 100644
--- a/src/audio_core/renderer/command/data_source/decode.cpp
+++ b/src/audio_core/renderer/command/data_source/decode.cpp
@@ -8,6 +8,7 @@
 #include "audio_core/renderer/command/resample/resample.h"
 #include "common/fixed_point.h"
 #include "common/logging/log.h"
+#include "common/scratch_buffer.h"
 #include "core/memory.h"
 namespace AudioCore::AudioRenderer {
@@ -27,6 +28,7 @@ constexpr std::array<u8, 3> PitchBySrcQuality = {4, 8, 4};
 template <typename T>
 static u32 DecodePcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
                     const DecodeArg& req) {
+    std::array<T, TempBufferSize> tmp_samples{};
    constexpr s32 min{std::numeric_limits<s16>::min()};
    constexpr s32 max{std::numeric_limits<s16>::max()};
@@ -49,18 +51,17 @@ static u32 DecodePcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
        const u64 size{channel_count * samples_to_decode};
        const u64 size_bytes{size * sizeof(T)};
-        std::vector<T> samples(size);
+        memory.ReadBlockUnsafe(source, tmp_samples.data(), size_bytes);
-        memory.ReadBlockUnsafe(source, samples.data(), size_bytes);
        if constexpr (std::is_floating_point_v<T>) {
            for (u32 i = 0; i < samples_to_decode; i++) {
-                auto sample{static_cast<s32>(samples[i * channel_count + req.target_channel] *
+                auto sample{static_cast<s32>(tmp_samples[i * channel_count + req.target_channel] *
                                             std::numeric_limits<s16>::max())};
                out_buffer[i] = static_cast<s16>(std::clamp(sample, min, max));
            }
        } else {
            for (u32 i = 0; i < samples_to_decode; i++) {
-                out_buffer[i] = samples[i * channel_count + req.target_channel];
+                out_buffer[i] = tmp_samples[i * channel_count + req.target_channel];
            }
        }
    } break;
@@ -73,17 +74,16 @@ static u32 DecodePcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
        }
        const VAddr source{req.buffer + ((req.start_offset + req.offset) * sizeof(T))};
-        std::vector<T> samples(samples_to_decode);
+        memory.ReadBlockUnsafe(source, tmp_samples.data(), samples_to_decode * sizeof(T));
-        memory.ReadBlockUnsafe(source, samples.data(), samples_to_decode * sizeof(T));
        if constexpr (std::is_floating_point_v<T>) {
            for (u32 i = 0; i < samples_to_decode; i++) {
-                auto sample{static_cast<s32>(samples[i * channel_count + req.target_channel] *
+                auto sample{static_cast<s32>(tmp_samples[i * channel_count + req.target_channel] *
                                             std::numeric_limits<s16>::max())};
                out_buffer[i] = static_cast<s16>(std::clamp(sample, min, max));
            }
        } else {
-            std::memcpy(out_buffer.data(), samples.data(), samples_to_decode * sizeof(s16));
+            std::memcpy(out_buffer.data(), tmp_samples.data(), samples_to_decode * sizeof(s16));
        }
        break;
    }
@@ -101,6 +101,7 @@ static u32 DecodePcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
 */
 static u32 DecodeAdpcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
                       const DecodeArg& req) {
+    std::array<u8, TempBufferSize> wavebuffer{};
    constexpr u32 SamplesPerFrame{14};
    constexpr u32 NibblesPerFrame{16};
@@ -138,9 +139,7 @@ static u32 DecodeAdpcm(Core::Memory::Memory& memory, std::span<s16> out_buffer,
    }
    const auto size{std::max((samples_to_process / 8U) * SamplesPerFrame, 8U)};
-    std::vector<u8> wavebuffer(size);
+    memory.ReadBlockUnsafe(req.buffer + position_in_frame / 2, wavebuffer.data(), size);
-    memory.ReadBlockUnsafe(req.buffer + position_in_frame / 2, wavebuffer.data(),
-                           wavebuffer.size());
    auto context{req.adpcm_context};
    auto header{context->header};
@@ -258,7 +257,7 @@ void DecodeFromWaveBuffers(Core::Memory::Memory& memory, const DecodeFromWaveBuf
    u32 offset{voice_state.offset};
    auto output_buffer{args.output};
-    std::vector<s16> temp_buffer(TempBufferSize, 0);
+    std::array<s16, TempBufferSize> temp_buffer{};
    while (remaining_sample_count > 0) {
        const auto samples_to_write{std::min(remaining_sample_count, max_remaining_sample_count)};
diff --git a/src/audio_core/renderer/command/effect/compressor.cpp b/src/audio_core/renderer/command/effect/compressor.cpp
index 7229618e8..ee9b68d5b 100644
--- a/src/audio_core/renderer/command/effect/compressor.cpp
+++ b/src/audio_core/renderer/command/effect/compressor.cpp
@@ -44,8 +44,8 @@ static void InitializeCompressorEffect(const CompressorInfo::ParameterVersion2&
 static void ApplyCompressorEffect(const CompressorInfo::ParameterVersion2& params,
                                  CompressorInfo::State& state, bool enabled,
-                                  std::vector<std::span<const s32>> input_buffers,
+                                  std::span<std::span<const s32>> input_buffers,
-                                  std::vector<std::span<s32>> output_buffers, u32 sample_count) {
+                                  std::span<std::span<s32>> output_buffers, u32 sample_count) {
    if (enabled) {
        auto state_00{state.unk_00};
        auto state_04{state.unk_04};
@@ -124,8 +124,8 @@ void CompressorCommand::Dump([[maybe_unused]] const ADSP::CommandListProcessor&
 }
 void CompressorCommand::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
+    std::array<std::span<const s32>, MaxChannels> input_buffers{};
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    std::array<std::span<s32>, MaxChannels> output_buffers{};
    for (s16 i = 0; i < parameter.channel_count; i++) {
        input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,
diff --git a/src/audio_core/renderer/command/effect/delay.cpp b/src/audio_core/renderer/command/effect/delay.cpp
index a4e408d40..e536cbb1e 100644
--- a/src/audio_core/renderer/command/effect/delay.cpp
+++ b/src/audio_core/renderer/command/effect/delay.cpp
@@ -51,7 +51,7 @@ static void InitializeDelayEffect(const DelayInfo::ParameterVersion1& params,
        state.delay_lines[channel].sample_count_max = sample_count_max.to_int_floor();
        state.delay_lines[channel].sample_count = sample_count.to_int_floor();
        state.delay_lines[channel].buffer.resize(state.delay_lines[channel].sample_count, 0);
-        if (state.delay_lines[channel].buffer.size() == 0) {
+        if (state.delay_lines[channel].sample_count == 0) {
            state.delay_lines[channel].buffer.push_back(0);
        }
        state.delay_lines[channel].buffer_pos = 0;
@@ -74,8 +74,8 @@ static void InitializeDelayEffect(const DelayInfo::ParameterVersion1& params,
 */
 template <size_t NumChannels>
 static void ApplyDelay(const DelayInfo::ParameterVersion1& params, DelayInfo::State& state,
-                       std::vector<std::span<const s32>>& inputs,
+                       std::span<std::span<const s32>> inputs, std::span<std::span<s32>> outputs,
-                       std::vector<std::span<s32>>& outputs, const u32 sample_count) {
+                       const u32 sample_count) {
    for (u32 sample_index = 0; sample_index < sample_count; sample_index++) {
        std::array<Common::FixedPoint<50, 14>, NumChannels> input_samples{};
        for (u32 channel = 0; channel < NumChannels; channel++) {
@@ -153,8 +153,8 @@ static void ApplyDelay(const DelayInfo::ParameterVersion1& params, DelayInfo::St
 * @param sample_count - Number of samples to process.
 */
 static void ApplyDelayEffect(const DelayInfo::ParameterVersion1& params, DelayInfo::State& state,
-                             const bool enabled, std::vector<std::span<const s32>>& inputs,
+                             const bool enabled, std::span<std::span<const s32>> inputs,
-                             std::vector<std::span<s32>>& outputs, const u32 sample_count) {
+                             std::span<std::span<s32>> outputs, const u32 sample_count) {
    if (!IsChannelCountValid(params.channel_count)) {
        LOG_ERROR(Service_Audio, "Invalid delay channels {}", params.channel_count);
@@ -208,8 +208,8 @@ void DelayCommand::Dump([[maybe_unused]] const ADSP::CommandListProcessor& proce
 }
 void DelayCommand::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
+    std::array<std::span<const s32>, MaxChannels> input_buffers{};
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    std::array<std::span<s32>, MaxChannels> output_buffers{};
    for (s16 i = 0; i < parameter.channel_count; i++) {
        input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,
diff --git a/src/audio_core/renderer/command/effect/i3dl2_reverb.cpp b/src/audio_core/renderer/command/effect/i3dl2_reverb.cpp
index 27d8b9844..d2bfb67cc 100644
--- a/src/audio_core/renderer/command/effect/i3dl2_reverb.cpp
+++ b/src/audio_core/renderer/command/effect/i3dl2_reverb.cpp
@@ -408,8 +408,8 @@ void I3dl2ReverbCommand::Dump([[maybe_unused]] const ADSP::CommandListProcessor&
 }
 void I3dl2ReverbCommand::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
+    std::array<std::span<const s32>, MaxChannels> input_buffers{};
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    std::array<std::span<s32>, MaxChannels> output_buffers{};
    for (u32 i = 0; i < parameter.channel_count; i++) {
        input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,
diff --git a/src/audio_core/renderer/command/effect/light_limiter.cpp b/src/audio_core/renderer/command/effect/light_limiter.cpp
index e8fb0e2fc..4161a9821 100644
--- a/src/audio_core/renderer/command/effect/light_limiter.cpp
+++ b/src/audio_core/renderer/command/effect/light_limiter.cpp
@@ -47,8 +47,8 @@ static void InitializeLightLimiterEffect(const LightLimiterInfo::ParameterVersio
 */
 static void ApplyLightLimiterEffect(const LightLimiterInfo::ParameterVersion2& params,
                                    LightLimiterInfo::State& state, const bool enabled,
-                                    std::vector<std::span<const s32>>& inputs,
+                                    std::span<std::span<const s32>> inputs,
-                                    std::vector<std::span<s32>>& outputs, const u32 sample_count,
+                                    std::span<std::span<s32>> outputs, const u32 sample_count,
                                    LightLimiterInfo::StatisticsInternal* statistics) {
    constexpr s64 min{std::numeric_limits<s32>::min()};
    constexpr s64 max{std::numeric_limits<s32>::max()};
@@ -147,8 +147,8 @@ void LightLimiterVersion1Command::Dump([[maybe_unused]] const ADSP::CommandListP
 }
 void LightLimiterVersion1Command::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
+    std::array<std::span<const s32>, MaxChannels> input_buffers{};
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    std::array<std::span<s32>, MaxChannels> output_buffers{};
    for (u32 i = 0; i < parameter.channel_count; i++) {
        input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,
@@ -190,8 +190,8 @@ void LightLimiterVersion2Command::Dump([[maybe_unused]] const ADSP::CommandListP
 }
 void LightLimiterVersion2Command::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
+    std::array<std::span<const s32>, MaxChannels> input_buffers{};
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    std::array<std::span<s32>, MaxChannels> output_buffers{};
    for (u32 i = 0; i < parameter.channel_count; i++) {
        input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,
diff --git a/src/audio_core/renderer/command/effect/reverb.cpp b/src/audio_core/renderer/command/effect/reverb.cpp
index 8b9b65214..fc2f15a5e 100644
--- a/src/audio_core/renderer/command/effect/reverb.cpp
+++ b/src/audio_core/renderer/command/effect/reverb.cpp
@@ -250,8 +250,8 @@ static Common::FixedPoint<50, 14> Axfx2AllPassTick(ReverbInfo::ReverbDelayLine&
 */
 template <size_t NumChannels>
 static void ApplyReverbEffect(const ReverbInfo::ParameterVersion2& params, ReverbInfo::State& state,
-                              std::vector<std::span<const s32>>& inputs,
+                              std::span<std::span<const s32>> inputs,
-                              std::vector<std::span<s32>>& outputs, const u32 sample_count) {
+                              std::span<std::span<s32>> outputs, const u32 sample_count) {
    static constexpr std::array<u8, ReverbInfo::MaxDelayTaps> OutTapIndexes1Ch{
        0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
    };
@@ -369,8 +369,8 @@ static void ApplyReverbEffect(const ReverbInfo::ParameterVersion2& params, Rever
 * @param sample_count - Number of samples to process.
 */
 static void ApplyReverbEffect(const ReverbInfo::ParameterVersion2& params, ReverbInfo::State& state,
-                              const bool enabled, std::vector<std::span<const s32>>& inputs,
+                              const bool enabled, std::span<std::span<const s32>> inputs,
-                              std::vector<std::span<s32>>& outputs, const u32 sample_count) {
+                              std::span<std::span<s32>> outputs, const u32 sample_count) {
    if (enabled) {
        switch (params.channel_count) {
        case 0:
@@ -412,8 +412,8 @@ void ReverbCommand::Dump([[maybe_unused]] const ADSP::CommandListProcessor& proc
 }
 void ReverbCommand::Process(const ADSP::CommandListProcessor& processor) {
-    std::vector<std::span<const s32>> input_buffers(parameter.channel_count);
+    std::array<std::span<const s32>, MaxChannels> input_buffers{};
-    std::vector<std::span<s32>> output_buffers(parameter.channel_count);
+    std::array<std::span<s32>, MaxChannels> output_buffers{};
    for (u32 i = 0; i < parameter.channel_count; i++) {
        input_buffers[i] = processor.mix_buffers.subspan(inputs[i] * processor.sample_count,
diff --git a/src/audio_core/renderer/command/performance/performance.cpp b/src/audio_core/renderer/command/performance/performance.cpp
index 985958b03..4a881547f 100644
--- a/src/audio_core/renderer/command/performance/performance.cpp
+++ b/src/audio_core/renderer/command/performance/performance.cpp
@@ -5,7 +5,6 @@
 #include "audio_core/renderer/command/performance/performance.h"
 #include "core/core.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 namespace AudioCore::AudioRenderer {
@@ -18,20 +17,18 @@ void PerformanceCommand::Process(const ADSP::CommandListProcessor& processor) {
    auto base{entry_address.translated_address};
    if (state == PerformanceState::Start) {
        auto start_time_ptr{reinterpret_cast<u32*>(base + entry_address.entry_start_time_offset)};
-        *start_time_ptr = static_cast<u32>(
+        *start_time_ptr =
-            Core::Timing::CyclesToUs(processor.system->CoreTiming().GetClockTicks() -
+            static_cast<u32>(processor.system->CoreTiming().GetClockTicks() - processor.start_time -
-                                     processor.start_time - processor.current_processing_time)
+                             processor.current_processing_time);
-                .count());
    } else if (state == PerformanceState::Stop) {
        auto processed_time_ptr{
            reinterpret_cast<u32*>(base + entry_address.entry_processed_time_offset)};
        auto entry_count_ptr{
            reinterpret_cast<u32*>(base + entry_address.header_entry_count_offset)};
-        *processed_time_ptr = static_cast<u32>(
+        *processed_time_ptr =
-            Core::Timing::CyclesToUs(processor.system->CoreTiming().GetClockTicks() -
+            static_cast<u32>(processor.system->CoreTiming().GetClockTicks() - processor.start_time -
-                                     processor.start_time - processor.current_processing_time)
+                             processor.current_processing_time);
-                .count());
        (*entry_count_ptr)++;
    }
 }
diff --git a/src/audio_core/renderer/command/sink/circular_buffer.cpp b/src/audio_core/renderer/command/sink/circular_buffer.cpp
index ded5afc94..e2ce59792 100644
--- a/src/audio_core/renderer/command/sink/circular_buffer.cpp
+++ b/src/audio_core/renderer/command/sink/circular_buffer.cpp
@@ -24,7 +24,7 @@ void CircularBufferSinkCommand::Process(const ADSP::CommandListProcessor& proces
    constexpr s32 min{std::numeric_limits<s16>::min()};
    constexpr s32 max{std::numeric_limits<s16>::max()};
-    std::vector<s16> output(processor.sample_count);
+    std::array<s16, TargetSampleCount * MaxChannels> output{};
    for (u32 channel = 0; channel < input_count; channel++) {
        auto input{processor.mix_buffers.subspan(inputs[channel] * processor.sample_count,
                                                 processor.sample_count)};
@@ -33,7 +33,7 @@ void CircularBufferSinkCommand::Process(const ADSP::CommandListProcessor& proces
        }
        processor.memory->WriteBlockUnsafe(address + pos, output.data(),
-                                           output.size() * sizeof(s16));
+                                           processor.sample_count * sizeof(s16));
        pos += static_cast<u32>(processor.sample_count * sizeof(s16));
        if (pos >= size) {
            pos = 0;
diff --git a/src/audio_core/renderer/command/sink/device.cpp b/src/audio_core/renderer/command/sink/device.cpp
index e88372a75..5f74dd7ad 100644
--- a/src/audio_core/renderer/command/sink/device.cpp
+++ b/src/audio_core/renderer/command/sink/device.cpp
@@ -33,8 +33,7 @@ void DeviceSinkCommand::Process(const ADSP::CommandListProcessor& processor) {
        .consumed{false},
    };
-    std::vector<s16> samples(out_buffer.frames * input_count);
+    std::array<s16, TargetSampleCount * MaxChannels> samples{};
    for (u32 channel = 0; channel < input_count; channel++) {
        const auto offset{inputs[channel] * out_buffer.frames};
@@ -45,7 +44,7 @@ void DeviceSinkCommand::Process(const ADSP::CommandListProcessor& processor) {
    }
    out_buffer.tag = reinterpret_cast<u64>(samples.data());
-    stream->AppendBuffer(out_buffer, samples);
+    stream->AppendBuffer(out_buffer, {samples.data(), out_buffer.frames * input_count});
    if (stream->IsPaused()) {
        stream->Start();
diff --git a/src/audio_core/renderer/mix/mix_context.cpp b/src/audio_core/renderer/mix/mix_context.cpp
index 35b748ede..3a18ae7c2 100644
--- a/src/audio_core/renderer/mix/mix_context.cpp
+++ b/src/audio_core/renderer/mix/mix_context.cpp
@@ -125,10 +125,10 @@ bool MixContext::TSortInfo(const SplitterContext& splitter_context) {
        return false;
    }
-    std::vector<s32> sorted_results{node_states.GetSortedResuls()};
+    auto sorted_results{node_states.GetSortedResuls()};
-    const auto result_size{std::min(count, static_cast<s32>(sorted_results.size()))};
+    const auto result_size{std::min(count, static_cast<s32>(sorted_results.second))};
    for (s32 i = 0; i < result_size; i++) {
-        sorted_mix_infos[i] = &mix_infos[sorted_results[i]];
+        sorted_mix_infos[i] = &mix_infos[sorted_results.first[i]];
    }
    CalcMixBufferOffset();
diff --git a/src/audio_core/renderer/nodes/node_states.cpp b/src/audio_core/renderer/nodes/node_states.cpp
index 1821a51e6..b7a44a54c 100644
--- a/src/audio_core/renderer/nodes/node_states.cpp
+++ b/src/audio_core/renderer/nodes/node_states.cpp
@@ -134,8 +134,8 @@ u32 NodeStates::GetNodeCount() const {
    return node_count;
 }
-std::vector<s32> NodeStates::GetSortedResuls() const {
+std::pair<std::span<u32>::reverse_iterator, size_t> NodeStates::GetSortedResuls() const {
-    return {results.rbegin(), results.rbegin() + result_pos};
+    return {results.rbegin(), result_pos};
 }
 } // namespace AudioCore::AudioRenderer
diff --git a/src/audio_core/renderer/nodes/node_states.h b/src/audio_core/renderer/nodes/node_states.h
index 94b1d1254..e768cd4b5 100644
--- a/src/audio_core/renderer/nodes/node_states.h
+++ b/src/audio_core/renderer/nodes/node_states.h
@@ -175,7 +175,7 @@ public:
     *
     * @return Vector of nodes in reverse order.
     */
-    std::vector<s32> GetSortedResuls() const;
+    std::pair<std::span<u32>::reverse_iterator, size_t> GetSortedResuls() const;
 private:
    /// Number of nodes in the graph
diff --git a/src/audio_core/renderer/system.cpp b/src/audio_core/renderer/system.cpp
index 53b258c4f..a23627472 100644
--- a/src/audio_core/renderer/system.cpp
+++ b/src/audio_core/renderer/system.cpp
@@ -444,6 +444,7 @@ Result System::Update(std::span<const u8> input, std::span<u8> performance, std:
    std::scoped_lock l{lock};
    const auto start_time{core.CoreTiming().GetClockTicks()};
+    std::memset(output.data(), 0, output.size());
    InfoUpdater info_updater(input, output, process_handle, behavior);
diff --git a/src/audio_core/sink/null_sink.h b/src/audio_core/sink/null_sink.h
index 1215d3cd2..b6b43c93e 100644
--- a/src/audio_core/sink/null_sink.h
+++ b/src/audio_core/sink/null_sink.h
@@ -20,7 +20,7 @@ public:
    explicit NullSinkStreamImpl(Core::System& system_, StreamType type_)
        : SinkStream{system_, type_} {}
    ~NullSinkStreamImpl() override {}
-    void AppendBuffer(SinkBuffer&, std::vector<s16>&) override {}
+    void AppendBuffer(SinkBuffer&, std::span<s16>) override {}
    std::vector<s16> ReleaseBuffer(u64) override {
        return {};
    }
diff --git a/src/audio_core/sink/sink_stream.cpp b/src/audio_core/sink/sink_stream.cpp
index f44fedfd5..404dcd0e9 100644
--- a/src/audio_core/sink/sink_stream.cpp
+++ b/src/audio_core/sink/sink_stream.cpp
@@ -15,11 +15,10 @@
 #include "common/settings.h"
 #include "core/core.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 namespace AudioCore::Sink {
-void SinkStream::AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples) {
+void SinkStream::AppendBuffer(SinkBuffer& buffer, std::span<s16> samples) {
    if (type == StreamType::In) {
        queue.enqueue(buffer);
        queued_buffers++;
@@ -67,15 +66,16 @@ void SinkStream::AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples) {
                static_cast<s16>(std::clamp(right_sample, min, max));
        }
-        samples.resize(samples.size() / system_channels * device_channels);
+        samples = samples.subspan(0, samples.size() / system_channels * device_channels);
    } else if (system_channels == 2 && device_channels == 6) {
        // We need moar samples! Not all games will provide 6 channel audio.
        // TODO: Implement some upmixing here. Currently just passthrough, with other
        // channels left as silence.
-        std::vector<s16> new_samples(samples.size() / system_channels * device_channels, 0);
+        auto new_size = samples.size() / system_channels * device_channels;
+        tmp_samples.resize_destructive(new_size);
-        for (u32 read_index = 0, write_index = 0; read_index < samples.size();
+        for (u32 read_index = 0, write_index = 0; read_index < new_size;
             read_index += system_channels, write_index += device_channels) {
            const auto left_sample{static_cast<s16>(std::clamp(
                static_cast<s32>(
@@ -83,7 +83,7 @@ void SinkStream::AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples) {
                    volume),
                min, max))};
-            new_samples[write_index + static_cast<u32>(Channels::FrontLeft)] = left_sample;
+            tmp_samples[write_index + static_cast<u32>(Channels::FrontLeft)] = left_sample;
            const auto right_sample{static_cast<s16>(std::clamp(
                static_cast<s32>(
@@ -91,9 +91,9 @@ void SinkStream::AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples) {
                    volume),
                min, max))};
-            new_samples[write_index + static_cast<u32>(Channels::FrontRight)] = right_sample;
+            tmp_samples[write_index + static_cast<u32>(Channels::FrontRight)] = right_sample;
        }
-        samples = std::move(new_samples);
+        samples = std::span<s16>(tmp_samples);
    } else if (volume != 1.0f) {
        for (u32 i = 0; i < samples.size(); i++) {
diff --git a/src/audio_core/sink/sink_stream.h b/src/audio_core/sink/sink_stream.h
index 41cbadc9c..98d72ace1 100644
--- a/src/audio_core/sink/sink_stream.h
+++ b/src/audio_core/sink/sink_stream.h
@@ -16,6 +16,7 @@
 #include "common/polyfill_thread.h"
 #include "common/reader_writer_queue.h"
 #include "common/ring_buffer.h"
+#include "common/scratch_buffer.h"
 #include "common/thread.h"
 namespace Core {
@@ -170,7 +171,7 @@ public:
     * @param buffer  - Audio buffer information to be queued.
     * @param samples - The s16 samples to be queue for playback.
     */
-    virtual void AppendBuffer(SinkBuffer& buffer, std::vector<s16>& samples);
+    virtual void AppendBuffer(SinkBuffer& buffer, std::span<s16> samples);
    /**
     * Release a buffer. Audio In only, will fill a buffer with recorded samples.
@@ -255,6 +256,8 @@ private:
    /// Signalled when ring buffer entries are consumed
    std::condition_variable_any release_cv;
    std::mutex release_mutex;
+    /// Temporary buffer for appending samples when upmixing
+    Common::ScratchBuffer<s16> tmp_samples{};
 };
 using SinkStreamPtr = std::unique_ptr<SinkStream>;
diff --git a/src/common/CMakeLists.txt b/src/common/CMakeLists.txt
index efc4a9fe9..3adf13a3f 100644
--- a/src/common/CMakeLists.txt
+++ b/src/common/CMakeLists.txt
@@ -172,6 +172,8 @@ if(ARCHITECTURE_x86_64)
            x64/cpu_wait.h
            x64/native_clock.cpp
            x64/native_clock.h
+            x64/rdtsc.cpp
+            x64/rdtsc.h
            x64/xbyak_abi.h
            x64/xbyak_util.h
    )
diff --git a/src/common/fs/fs.cpp b/src/common/fs/fs.cpp
index 6d66c926d..1baf6d746 100644
--- a/src/common/fs/fs.cpp
+++ b/src/common/fs/fs.cpp
@@ -436,7 +436,7 @@ void IterateDirEntries(const std::filesystem::path& path, const DirEntryCallable
        if (True(filter & DirEntryFilter::File) &&
            entry.status().type() == fs::file_type::regular) {
-            if (!callback(entry.path())) {
+            if (!callback(entry)) {
                callback_error = true;
                break;
            }
@@ -444,7 +444,7 @@ void IterateDirEntries(const std::filesystem::path& path, const DirEntryCallable
        if (True(filter & DirEntryFilter::Directory) &&
            entry.status().type() == fs::file_type::directory) {
-            if (!callback(entry.path())) {
+            if (!callback(entry)) {
                callback_error = true;
                break;
            }
@@ -493,7 +493,7 @@ void IterateDirEntriesRecursively(const std::filesystem::path& path,
        if (True(filter & DirEntryFilter::File) &&
            entry.status().type() == fs::file_type::regular) {
-            if (!callback(entry.path())) {
+            if (!callback(entry)) {
                callback_error = true;
                break;
            }
@@ -501,7 +501,7 @@ void IterateDirEntriesRecursively(const std::filesystem::path& path,
        if (True(filter & DirEntryFilter::Directory) &&
            entry.status().type() == fs::file_type::directory) {
-            if (!callback(entry.path())) {
+            if (!callback(entry)) {
                callback_error = true;
                break;
            }
diff --git a/src/common/fs/fs_types.h b/src/common/fs/fs_types.h
index 5a4090c19..900f85d24 100644
--- a/src/common/fs/fs_types.h
+++ b/src/common/fs/fs_types.h
@@ -66,6 +66,6 @@ DECLARE_ENUM_FLAG_OPERATORS(DirEntryFilter);
 * @returns A boolean value.
 *          Return true to indicate whether the callback is successful, false otherwise.
 */
-using DirEntryCallable = std::function<bool(const std::filesystem::path& path)>;
+using DirEntryCallable = std::function<bool(const std::filesystem::directory_entry& entry)>;
 } // namespace Common::FS
diff --git a/src/common/input.h b/src/common/input.h
index 66fb15f0a..ea30770ae 100644
--- a/src/common/input.h
+++ b/src/common/input.h
@@ -86,7 +86,7 @@ enum class NfcState {
    NewAmiibo,
    WaitingForAmiibo,
    AmiiboRemoved,
-    NotAnAmiibo,
+    InvalidTagType,
    NotSupported,
    WrongDeviceState,
    WriteFailed,
@@ -218,8 +218,22 @@ struct CameraStatus {
 };
 struct NfcStatus {
-    NfcState state{};
+    NfcState state{NfcState::Unknown};
-    std::vector<u8> data{};
+    u8 uuid_length;
+    u8 protocol;
+    u8 tag_type;
+    std::array<u8, 10> uuid;
+};
+struct MifareData {
+    u8 command;
+    u8 sector;
+    std::array<u8, 0x6> key;
+    std::array<u8, 0x10> data;
+};
+struct MifareRequest {
+    std::array<MifareData, 0x10> data;
 };
 // List of buttons to be passed to Qt that can be translated
@@ -294,7 +308,7 @@ struct CallbackStatus {
    BatteryStatus battery_status{};
    VibrationStatus vibration_status{};
    CameraFormat camera_status{CameraFormat::None};
-    NfcState nfc_status{NfcState::Unknown};
+    NfcStatus nfc_status{};
    std::vector<u8> raw_data{};
 };
@@ -356,9 +370,30 @@ public:
        return NfcState::NotSupported;
    }
+    virtual NfcState StartNfcPolling() {
+        return NfcState::NotSupported;
+    }
+    virtual NfcState StopNfcPolling() {
+        return NfcState::NotSupported;
+    }
+    virtual NfcState ReadAmiiboData([[maybe_unused]] std::vector<u8>& out_data) {
+        return NfcState::NotSupported;
+    }
    virtual NfcState WriteNfcData([[maybe_unused]] const std::vector<u8>& data) {
        return NfcState::NotSupported;
    }
+    virtual NfcState ReadMifareData([[maybe_unused]] const MifareRequest& request,
+                                    [[maybe_unused]] MifareRequest& out_data) {
+        return NfcState::NotSupported;
+    }
+    virtual NfcState WriteMifareData([[maybe_unused]] const MifareRequest& request) {
+        return NfcState::NotSupported;
+    }
 };
 /// An abstract class template for a factory that can create input devices.
diff --git a/src/common/ring_buffer.h b/src/common/ring_buffer.h
index 4c328ab44..416680d44 100644
--- a/src/common/ring_buffer.h
+++ b/src/common/ring_buffer.h
@@ -9,6 +9,7 @@
 #include <cstddef>
 #include <cstring>
 #include <new>
+#include <span>
 #include <type_traits>
 #include <vector>
@@ -53,7 +54,7 @@ public:
        return push_count;
    }
-    std::size_t Push(const std::vector<T>& input) {
+    std::size_t Push(const std::span<T> input) {
        return Push(input.data(), input.size());
    }
diff --git a/src/common/scratch_buffer.h b/src/common/scratch_buffer.h
index a69a5a7af..6fe907953 100644
--- a/src/common/scratch_buffer.h
+++ b/src/common/scratch_buffer.h
@@ -3,6 +3,9 @@
 #pragma once
+#include <iterator>
+#include "common/concepts.h"
 #include "common/make_unique_for_overwrite.h"
 namespace Common {
@@ -16,6 +19,12 @@ namespace Common {
 template <typename T>
 class ScratchBuffer {
 public:
+    using iterator = T*;
+    using const_iterator = const T*;
+    using value_type = T;
+    using element_type = T;
+    using iterator_category = std::contiguous_iterator_tag;
    ScratchBuffer() = default;
    explicit ScratchBuffer(size_t initial_capacity)
diff --git a/src/common/settings.h b/src/common/settings.h
index 9682281b0..3aedf3850 100644
--- a/src/common/settings.h
+++ b/src/common/settings.h
@@ -483,6 +483,7 @@ struct Values {
        AstcRecompression::Uncompressed, AstcRecompression::Uncompressed, AstcRecompression::Bc3,
        "astc_recompression"};
    SwitchableSetting<bool> use_video_framerate{false, "use_video_framerate"};
+    SwitchableSetting<bool> barrier_feedback_loops{true, "barrier_feedback_loops"};
    SwitchableSetting<u8> bg_red{0, "bg_red"};
    SwitchableSetting<u8> bg_green{0, "bg_green"};
diff --git a/src/common/steady_clock.cpp b/src/common/steady_clock.cpp
index 782859196..9415eed29 100644
--- a/src/common/steady_clock.cpp
+++ b/src/common/steady_clock.cpp
@@ -28,13 +28,12 @@ static s64 GetSystemTimeNS() {
    // GetSystemTimePreciseAsFileTime returns the file time in 100ns units.
    static constexpr s64 Multiplier = 100;
    // Convert Windows epoch to Unix epoch.
-    static constexpr s64 WindowsEpochToUnixEpochNS = 0x19DB1DED53E8000LL;
+    static constexpr s64 WindowsEpochToUnixEpoch = 0x19DB1DED53E8000LL;
    FILETIME filetime;
    GetSystemTimePreciseAsFileTime(&filetime);
    return Multiplier * ((static_cast<s64>(filetime.dwHighDateTime) << 32) +
-                         static_cast<s64>(filetime.dwLowDateTime)) -
+                         static_cast<s64>(filetime.dwLowDateTime) - WindowsEpochToUnixEpoch);
-           WindowsEpochToUnixEpochNS;
 }
 #endif
diff --git a/src/common/thread.h b/src/common/thread.h
index 8ae169b4e..c6976fb6c 100644
--- a/src/common/thread.h
+++ b/src/common/thread.h
@@ -55,7 +55,7 @@ public:
        is_set = false;
    }
-    [[nodiscard]] bool IsSet() {
+    [[nodiscard]] bool IsSet() const {
        return is_set;
    }
diff --git a/src/common/wall_clock.cpp b/src/common/wall_clock.cpp
index 817e71d52..dc0dcbd68 100644
--- a/src/common/wall_clock.cpp
+++ b/src/common/wall_clock.cpp
@@ -2,88 +2,75 @@
 // SPDX-License-Identifier: GPL-2.0-or-later
 #include "common/steady_clock.h"
-#include "common/uint128.h"
 #include "common/wall_clock.h"
 #ifdef ARCHITECTURE_x86_64
 #include "common/x64/cpu_detect.h"
 #include "common/x64/native_clock.h"
+#include "common/x64/rdtsc.h"
 #endif
 namespace Common {
 class StandardWallClock final : public WallClock {
 public:
-    explicit StandardWallClock(u64 emulated_cpu_frequency_, u64 emulated_clock_frequency_)
+    explicit StandardWallClock() : start_time{SteadyClock::Now()} {}
-        : WallClock{emulated_cpu_frequency_, emulated_clock_frequency_, false},
-          start_time{SteadyClock::Now()} {}
-    std::chrono::nanoseconds GetTimeNS() override {
+    std::chrono::nanoseconds GetTimeNS() const override {
        return SteadyClock::Now() - start_time;
    }
-    std::chrono::microseconds GetTimeUS() override {
+    std::chrono::microseconds GetTimeUS() const override {
-        return std::chrono::duration_cast<std::chrono::microseconds>(GetTimeNS());
+        return static_cast<std::chrono::microseconds>(GetHostTicksElapsed() / NsToUsRatio::den);
    }
-    std::chrono::milliseconds GetTimeMS() override {
+    std::chrono::milliseconds GetTimeMS() const override {
-        return std::chrono::duration_cast<std::chrono::milliseconds>(GetTimeNS());
+        return static_cast<std::chrono::milliseconds>(GetHostTicksElapsed() / NsToMsRatio::den);
    }
-    u64 GetClockCycles() override {
+    u64 GetCNTPCT() const override {
-        const u128 temp = Common::Multiply64Into128(GetTimeNS().count(), emulated_clock_frequency);
+        return GetHostTicksElapsed() * NsToCNTPCTRatio::num / NsToCNTPCTRatio::den;
-        return Common::Divide128On32(temp, NS_RATIO).first;
    }
-    u64 GetCPUCycles() override {
+    u64 GetGPUTick() const override {
-        const u128 temp = Common::Multiply64Into128(GetTimeNS().count(), emulated_cpu_frequency);
+        return GetHostTicksElapsed() * NsToGPUTickRatio::num / NsToGPUTickRatio::den;
-        return Common::Divide128On32(temp, NS_RATIO).first;
    }
-    void Pause([[maybe_unused]] bool is_paused) override {
+    u64 GetHostTicksNow() const override {
-        // Do nothing in this clock type.
+        return static_cast<u64>(SteadyClock::Now().time_since_epoch().count());
+    }
+    u64 GetHostTicksElapsed() const override {
+        return static_cast<u64>(GetTimeNS().count());
+    }
+    bool IsNative() const override {
+        return false;
    }
 private:
    SteadyClock::time_point start_time;
 };
+std::unique_ptr<WallClock> CreateOptimalClock() {
 #ifdef ARCHITECTURE_x86_64
-std::unique_ptr<WallClock> CreateBestMatchingClock(u64 emulated_cpu_frequency,
-                                                   u64 emulated_clock_frequency) {
    const auto& caps = GetCPUCaps();
-    u64 rtsc_frequency = 0;
-    if (caps.invariant_tsc) {
-        rtsc_frequency = caps.tsc_frequency ? caps.tsc_frequency : EstimateRDTSCFrequency();
-    }
-    // Fallback to StandardWallClock if the hardware TSC does not have the precision greater than:
+    if (caps.invariant_tsc && caps.tsc_frequency >= WallClock::GPUTickFreq) {
-    // - A nanosecond
+        return std::make_unique<X64::NativeClock>(caps.tsc_frequency);
-    // - The emulated CPU frequency
-    // - The emulated clock counter frequency (CNTFRQ)
-    if (rtsc_frequency <= WallClock::NS_RATIO || rtsc_frequency <= emulated_cpu_frequency ||
-        rtsc_frequency <= emulated_clock_frequency) {
-        return std::make_unique<StandardWallClock>(emulated_cpu_frequency,
-                                                   emulated_clock_frequency);
    } else {
-        return std::make_unique<X64::NativeClock>(emulated_cpu_frequency, emulated_clock_frequency,
+        // Fallback to StandardWallClock if the hardware TSC
-                                                  rtsc_frequency);
+        // - Is not invariant
+        // - Is not more precise than GPUTickFreq
+        return std::make_unique<StandardWallClock>();
    }
-}
 #else
+    return std::make_unique<StandardWallClock>();
-std::unique_ptr<WallClock> CreateBestMatchingClock(u64 emulated_cpu_frequency,
-                                                   u64 emulated_clock_frequency) {
-    return std::make_unique<StandardWallClock>(emulated_cpu_frequency, emulated_clock_frequency);
-}
 #endif
+}
-std::unique_ptr<WallClock> CreateStandardWallClock(u64 emulated_cpu_frequency,
+std::unique_ptr<WallClock> CreateStandardWallClock() {
-                                                   u64 emulated_clock_frequency) {
+    return std::make_unique<StandardWallClock>();
-    return std::make_unique<StandardWallClock>(emulated_cpu_frequency, emulated_clock_frequency);
 }
 } // namespace Common
diff --git a/src/common/wall_clock.h b/src/common/wall_clock.h
index 157ec5eae..f45d3d8c5 100644
--- a/src/common/wall_clock.h
+++ b/src/common/wall_clock.h
@@ -5,6 +5,7 @@
 #include <chrono>
 #include <memory>
+#include <ratio>
 #include "common/common_types.h"
@@ -12,50 +13,82 @@ namespace Common {
 class WallClock {
 public:
-    static constexpr u64 NS_RATIO = 1'000'000'000;
+    static constexpr u64 CNTFRQ = 19'200'000;         // CNTPCT_EL0 Frequency = 19.2 MHz
-    static constexpr u64 US_RATIO = 1'000'000;
+    static constexpr u64 GPUTickFreq = 614'400'000;   // GM20B GPU Tick Frequency = 614.4 MHz
-    static constexpr u64 MS_RATIO = 1'000;
+    static constexpr u64 CPUTickFreq = 1'020'000'000; // T210/4 A57 CPU Tick Frequency = 1020.0 MHz
    virtual ~WallClock() = default;
-    /// Returns current wall time in nanoseconds
+    /// @returns The time in nanoseconds since the construction of this clock.
-    [[nodiscard]] virtual std::chrono::nanoseconds GetTimeNS() = 0;
+    virtual std::chrono::nanoseconds GetTimeNS() const = 0;
-    /// Returns current wall time in microseconds
+    /// @returns The time in microseconds since the construction of this clock.
-    [[nodiscard]] virtual std::chrono::microseconds GetTimeUS() = 0;
+    virtual std::chrono::microseconds GetTimeUS() const = 0;
-    /// Returns current wall time in milliseconds
+    /// @returns The time in milliseconds since the construction of this clock.
-    [[nodiscard]] virtual std::chrono::milliseconds GetTimeMS() = 0;
+    virtual std::chrono::milliseconds GetTimeMS() const = 0;
-    /// Returns current wall time in emulated clock cycles
+    /// @returns The guest CNTPCT ticks since the construction of this clock.
-    [[nodiscard]] virtual u64 GetClockCycles() = 0;
+    virtual u64 GetCNTPCT() const = 0;
-    /// Returns current wall time in emulated cpu cycles
+    /// @returns The guest GPU ticks since the construction of this clock.
-    [[nodiscard]] virtual u64 GetCPUCycles() = 0;
+    virtual u64 GetGPUTick() const = 0;
-    virtual void Pause(bool is_paused) = 0;
+    /// @returns The raw host timer ticks since an indeterminate epoch.
+    virtual u64 GetHostTicksNow() const = 0;
-    /// Tells if the wall clock, uses the host CPU's hardware clock
+    /// @returns The raw host timer ticks since the construction of this clock.
-    [[nodiscard]] bool IsNative() const {
+    virtual u64 GetHostTicksElapsed() const = 0;
-        return is_native;
+    /// @returns Whether the clock directly uses the host's hardware clock.
+    virtual bool IsNative() const = 0;
+    static inline u64 NSToCNTPCT(u64 ns) {
+        return ns * NsToCNTPCTRatio::num / NsToCNTPCTRatio::den;
+    }
+    static inline u64 NSToGPUTick(u64 ns) {
+        return ns * NsToGPUTickRatio::num / NsToGPUTickRatio::den;
+    }
+    // Cycle Timing
+    static inline u64 CPUTickToNS(u64 cpu_tick) {
+        return cpu_tick * CPUTickToNsRatio::num / CPUTickToNsRatio::den;
+    }
+    static inline u64 CPUTickToUS(u64 cpu_tick) {
+        return cpu_tick * CPUTickToUsRatio::num / CPUTickToUsRatio::den;
+    }
+    static inline u64 CPUTickToCNTPCT(u64 cpu_tick) {
+        return cpu_tick * CPUTickToCNTPCTRatio::num / CPUTickToCNTPCTRatio::den;
+    }
+    static inline u64 CPUTickToGPUTick(u64 cpu_tick) {
+        return cpu_tick * CPUTickToGPUTickRatio::num / CPUTickToGPUTickRatio::den;
    }
 protected:
-    explicit WallClock(u64 emulated_cpu_frequency_, u64 emulated_clock_frequency_, bool is_native_)
+    using NsRatio = std::nano;
-        : emulated_cpu_frequency{emulated_cpu_frequency_},
+    using UsRatio = std::micro;
-          emulated_clock_frequency{emulated_clock_frequency_}, is_native{is_native_} {}
+    using MsRatio = std::milli;
+    using NsToUsRatio = std::ratio_divide<std::nano, std::micro>;
+    using NsToMsRatio = std::ratio_divide<std::nano, std::milli>;
+    using NsToCNTPCTRatio = std::ratio<CNTFRQ, std::nano::den>;
+    using NsToGPUTickRatio = std::ratio<GPUTickFreq, std::nano::den>;
-    u64 emulated_cpu_frequency;
+    // Cycle Timing
-    u64 emulated_clock_frequency;
-private:
+    using CPUTickToNsRatio = std::ratio<std::nano::den, CPUTickFreq>;
-    bool is_native;
+    using CPUTickToUsRatio = std::ratio<std::micro::den, CPUTickFreq>;
+    using CPUTickToCNTPCTRatio = std::ratio<CNTFRQ, CPUTickFreq>;
+    using CPUTickToGPUTickRatio = std::ratio<GPUTickFreq, CPUTickFreq>;
 };
-[[nodiscard]] std::unique_ptr<WallClock> CreateBestMatchingClock(u64 emulated_cpu_frequency,
+std::unique_ptr<WallClock> CreateOptimalClock();
-                                                                 u64 emulated_clock_frequency);
-[[nodiscard]] std::unique_ptr<WallClock> CreateStandardWallClock(u64 emulated_cpu_frequency,
+std::unique_ptr<WallClock> CreateStandardWallClock();
-                                                                 u64 emulated_clock_frequency);
 } // namespace Common
diff --git a/src/common/x64/cpu_detect.cpp b/src/common/x64/cpu_detect.cpp
index 72ed6e96c..c998b1197 100644
--- a/src/common/x64/cpu_detect.cpp
+++ b/src/common/x64/cpu_detect.cpp
@@ -14,6 +14,7 @@
 #include "common/common_types.h"
 #include "common/logging/log.h"
 #include "common/x64/cpu_detect.h"
+#include "common/x64/rdtsc.h"
 #ifdef _WIN32
 #include <windows.h>
@@ -187,6 +188,8 @@ static CPUCaps Detect() {
            caps.tsc_frequency = static_cast<u64>(caps.crystal_frequency) *
                                 caps.tsc_crystal_ratio_numerator /
                                 caps.tsc_crystal_ratio_denominator;
+        } else {
+            caps.tsc_frequency = X64::EstimateRDTSCFrequency();
        }
    }
diff --git a/src/common/x64/cpu_wait.cpp b/src/common/x64/cpu_wait.cpp
index cfeef6a3d..c53dd4945 100644
--- a/src/common/x64/cpu_wait.cpp
+++ b/src/common/x64/cpu_wait.cpp
@@ -9,19 +9,11 @@
 #include "common/x64/cpu_detect.h"
 #include "common/x64/cpu_wait.h"
+#include "common/x64/rdtsc.h"
 namespace Common::X64 {
 #ifdef _MSC_VER
-__forceinline static u64 FencedRDTSC() {
-    _mm_lfence();
-    _ReadWriteBarrier();
-    const u64 result = __rdtsc();
-    _mm_lfence();
-    _ReadWriteBarrier();
-    return result;
-}
 __forceinline static void TPAUSE() {
    // 100,000 cycles is a reasonable amount of time to wait to save on CPU resources.
    // For reference:
@@ -32,16 +24,6 @@ __forceinline static void TPAUSE() {
    _tpause(0, FencedRDTSC() + PauseCycles);
 }
 #else
-static u64 FencedRDTSC() {
-    u64 eax;
-    u64 edx;
-    asm volatile("lfence\n\t"
-                 "rdtsc\n\t"
-                 "lfence\n\t"
-                 : "=a"(eax), "=d"(edx));
-    return (edx << 32) | eax;
-}
 static void TPAUSE() {
    // 100,000 cycles is a reasonable amount of time to wait to save on CPU resources.
    // For reference:
diff --git a/src/common/x64/native_clock.cpp b/src/common/x64/native_clock.cpp
index 277b00662..7d2a26bd9 100644
--- a/src/common/x64/native_clock.cpp
+++ b/src/common/x64/native_clock.cpp
@@ -1,164 +1,50 @@
 // SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-2.0-or-later
-#include <array>
-#include <chrono>
-#include <thread>
-#include "common/atomic_ops.h"
-#include "common/steady_clock.h"
 #include "common/uint128.h"
 #include "common/x64/native_clock.h"
+#include "common/x64/rdtsc.h"
-#ifdef _MSC_VER
+namespace Common::X64 {
-#include <intrin.h>
-#endif
-namespace Common {
-#ifdef _MSC_VER
+NativeClock::NativeClock(u64 rdtsc_frequency_)
-__forceinline static u64 FencedRDTSC() {
+    : start_ticks{FencedRDTSC()}, rdtsc_frequency{rdtsc_frequency_},
-    _mm_lfence();
+      ns_rdtsc_factor{GetFixedPoint64Factor(NsRatio::den, rdtsc_frequency)},
-    _ReadWriteBarrier();
+      us_rdtsc_factor{GetFixedPoint64Factor(UsRatio::den, rdtsc_frequency)},
-    const u64 result = __rdtsc();
+      ms_rdtsc_factor{GetFixedPoint64Factor(MsRatio::den, rdtsc_frequency)},
-    _mm_lfence();
+      cntpct_rdtsc_factor{GetFixedPoint64Factor(CNTFRQ, rdtsc_frequency)},
-    _ReadWriteBarrier();
+      gputick_rdtsc_factor{GetFixedPoint64Factor(GPUTickFreq, rdtsc_frequency)} {}
-    return result;
-}
-#else
-static u64 FencedRDTSC() {
-    u64 eax;
-    u64 edx;
-    asm volatile("lfence\n\t"
-                 "rdtsc\n\t"
-                 "lfence\n\t"
-                 : "=a"(eax), "=d"(edx));
-    return (edx << 32) | eax;
-}
-#endif
-template <u64 Nearest>
+std::chrono::nanoseconds NativeClock::GetTimeNS() const {
-static u64 RoundToNearest(u64 value) {
+    return std::chrono::nanoseconds{MultiplyHigh(GetHostTicksElapsed(), ns_rdtsc_factor)};
-    const auto mod = value % Nearest;
-    return mod >= (Nearest / 2) ? (value - mod + Nearest) : (value - mod);
 }
-u64 EstimateRDTSCFrequency() {
+std::chrono::microseconds NativeClock::GetTimeUS() const {
-    // Discard the first result measuring the rdtsc.
+    return std::chrono::microseconds{MultiplyHigh(GetHostTicksElapsed(), us_rdtsc_factor)};
-    FencedRDTSC();
-    std::this_thread::sleep_for(std::chrono::milliseconds{1});
-    FencedRDTSC();
-    // Get the current time.
-    const auto start_time = Common::RealTimeClock::Now();
-    const u64 tsc_start = FencedRDTSC();
-    // Wait for 250 milliseconds.
-    std::this_thread::sleep_for(std::chrono::milliseconds{250});
-    const auto end_time = Common::RealTimeClock::Now();
-    const u64 tsc_end = FencedRDTSC();
-    // Calculate differences.
-    const u64 timer_diff = static_cast<u64>(
-        std::chrono::duration_cast<std::chrono::nanoseconds>(end_time - start_time).count());
-    const u64 tsc_diff = tsc_end - tsc_start;
-    const u64 tsc_freq = MultiplyAndDivide64(tsc_diff, 1000000000ULL, timer_diff);
-    return RoundToNearest<1000>(tsc_freq);
 }
-namespace X64 {
+std::chrono::milliseconds NativeClock::GetTimeMS() const {
-NativeClock::NativeClock(u64 emulated_cpu_frequency_, u64 emulated_clock_frequency_,
+    return std::chrono::milliseconds{MultiplyHigh(GetHostTicksElapsed(), ms_rdtsc_factor)};
-                         u64 rtsc_frequency_)
-    : WallClock(emulated_cpu_frequency_, emulated_clock_frequency_, true), rtsc_frequency{
-                                                                               rtsc_frequency_} {
-    // Thread to re-adjust the RDTSC frequency after 10 seconds has elapsed.
-    time_sync_thread = std::jthread{[this](std::stop_token token) {
-        // Get the current time.
-        const auto start_time = Common::RealTimeClock::Now();
-        const u64 tsc_start = FencedRDTSC();
-        // Wait for 10 seconds.
-        if (!Common::StoppableTimedWait(token, std::chrono::seconds{10})) {
-            return;
-        }
-        const auto end_time = Common::RealTimeClock::Now();
-        const u64 tsc_end = FencedRDTSC();
-        // Calculate differences.
-        const u64 timer_diff = static_cast<u64>(
-            std::chrono::duration_cast<std::chrono::nanoseconds>(end_time - start_time).count());
-        const u64 tsc_diff = tsc_end - tsc_start;
-        const u64 tsc_freq = MultiplyAndDivide64(tsc_diff, 1000000000ULL, timer_diff);
-        rtsc_frequency = tsc_freq;
-        CalculateAndSetFactors();
-    }};
-    time_point.inner.last_measure = FencedRDTSC();
-    time_point.inner.accumulated_ticks = 0U;
-    CalculateAndSetFactors();
 }
-u64 NativeClock::GetRTSC() {
+u64 NativeClock::GetCNTPCT() const {
-    TimePoint new_time_point{};
+    return MultiplyHigh(GetHostTicksElapsed(), cntpct_rdtsc_factor);
-    TimePoint current_time_point{};
-    current_time_point.pack = Common::AtomicLoad128(time_point.pack.data());
-    do {
-        const u64 current_measure = FencedRDTSC();
-        u64 diff = current_measure - current_time_point.inner.last_measure;
-        diff = diff & ~static_cast<u64>(static_cast<s64>(diff) >> 63); // max(diff, 0)
-        new_time_point.inner.last_measure = current_measure > current_time_point.inner.last_measure
-                                                ? current_measure
-                                                : current_time_point.inner.last_measure;
-        new_time_point.inner.accumulated_ticks = current_time_point.inner.accumulated_ticks + diff;
-    } while (!Common::AtomicCompareAndSwap(time_point.pack.data(), new_time_point.pack,
-                                           current_time_point.pack, current_time_point.pack));
-    return new_time_point.inner.accumulated_ticks;
 }
-void NativeClock::Pause(bool is_paused) {
+u64 NativeClock::GetGPUTick() const {
-    if (!is_paused) {
+    return MultiplyHigh(GetHostTicksElapsed(), gputick_rdtsc_factor);
-        TimePoint current_time_point{};
-        TimePoint new_time_point{};
-        current_time_point.pack = Common::AtomicLoad128(time_point.pack.data());
-        do {
-            new_time_point.pack = current_time_point.pack;
-            new_time_point.inner.last_measure = FencedRDTSC();
-        } while (!Common::AtomicCompareAndSwap(time_point.pack.data(), new_time_point.pack,
-                                               current_time_point.pack, current_time_point.pack));
-    }
 }
-std::chrono::nanoseconds NativeClock::GetTimeNS() {
+u64 NativeClock::GetHostTicksNow() const {
-    const u64 rtsc_value = GetRTSC();
+    return FencedRDTSC();
-    return std::chrono::nanoseconds{MultiplyHigh(rtsc_value, ns_rtsc_factor)};
 }
-std::chrono::microseconds NativeClock::GetTimeUS() {
+u64 NativeClock::GetHostTicksElapsed() const {
-    const u64 rtsc_value = GetRTSC();
+    return FencedRDTSC() - start_ticks;
-    return std::chrono::microseconds{MultiplyHigh(rtsc_value, us_rtsc_factor)};
 }
-std::chrono::milliseconds NativeClock::GetTimeMS() {
+bool NativeClock::IsNative() const {
-    const u64 rtsc_value = GetRTSC();
+    return true;
-    return std::chrono::milliseconds{MultiplyHigh(rtsc_value, ms_rtsc_factor)};
 }
-u64 NativeClock::GetClockCycles() {
+} // namespace Common::X64
-    const u64 rtsc_value = GetRTSC();
-    return MultiplyHigh(rtsc_value, clock_rtsc_factor);
-}
-u64 NativeClock::GetCPUCycles() {
-    const u64 rtsc_value = GetRTSC();
-    return MultiplyHigh(rtsc_value, cpu_rtsc_factor);
-}
-void NativeClock::CalculateAndSetFactors() {
-    ns_rtsc_factor = GetFixedPoint64Factor(NS_RATIO, rtsc_frequency);
-    us_rtsc_factor = GetFixedPoint64Factor(US_RATIO, rtsc_frequency);
-    ms_rtsc_factor = GetFixedPoint64Factor(MS_RATIO, rtsc_frequency);
-    clock_rtsc_factor = GetFixedPoint64Factor(emulated_clock_frequency, rtsc_frequency);
-    cpu_rtsc_factor = GetFixedPoint64Factor(emulated_cpu_frequency, rtsc_frequency);
-}
-} // namespace X64
-} // namespace Common
diff --git a/src/common/x64/native_clock.h b/src/common/x64/native_clock.h
index 03ca291d8..334415eff 100644
--- a/src/common/x64/native_clock.h
+++ b/src/common/x64/native_clock.h
@@ -3,58 +3,39 @@
 #pragma once
-#include "common/polyfill_thread.h"
 #include "common/wall_clock.h"
-namespace Common {
+namespace Common::X64 {
-namespace X64 {
 class NativeClock final : public WallClock {
 public:
-    explicit NativeClock(u64 emulated_cpu_frequency_, u64 emulated_clock_frequency_,
+    explicit NativeClock(u64 rdtsc_frequency_);
-                         u64 rtsc_frequency_);
-    std::chrono::nanoseconds GetTimeNS() override;
+    std::chrono::nanoseconds GetTimeNS() const override;
-    std::chrono::microseconds GetTimeUS() override;
+    std::chrono::microseconds GetTimeUS() const override;
-    std::chrono::milliseconds GetTimeMS() override;
+    std::chrono::milliseconds GetTimeMS() const override;
-    u64 GetClockCycles() override;
+    u64 GetCNTPCT() const override;
-    u64 GetCPUCycles() override;
+    u64 GetGPUTick() const override;
-    void Pause(bool is_paused) override;
+    u64 GetHostTicksNow() const override;
-private:
+    u64 GetHostTicksElapsed() const override;
-    u64 GetRTSC();
-    void CalculateAndSetFactors();
-    union alignas(16) TimePoint {
-        TimePoint() : pack{} {}
-        u128 pack{};
-        struct Inner {
-            u64 last_measure{};
-            u64 accumulated_ticks{};
-        } inner;
-    };
-    TimePoint time_point;
-    // factors
+    bool IsNative() const override;
-    u64 clock_rtsc_factor{};
-    u64 cpu_rtsc_factor{};
-    u64 ns_rtsc_factor{};
-    u64 us_rtsc_factor{};
-    u64 ms_rtsc_factor{};
-    u64 rtsc_frequency;
+private:
+    u64 start_ticks;
-    std::jthread time_sync_thread;
+    u64 rdtsc_frequency;
+    u64 ns_rdtsc_factor;
+    u64 us_rdtsc_factor;
+    u64 ms_rdtsc_factor;
+    u64 cntpct_rdtsc_factor;
+    u64 gputick_rdtsc_factor;
 };
-} // namespace X64
-u64 EstimateRDTSCFrequency();
-} // namespace Common
+} // namespace Common::X64
diff --git a/src/common/x64/rdtsc.cpp b/src/common/x64/rdtsc.cpp
new file mode 100644
index 000000000..9273274a3
--- /dev/null
+++ b/src/common/x64/rdtsc.cpp
@@ -0,0 +1,39 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+#include <thread>
+#include "common/steady_clock.h"
+#include "common/uint128.h"
+#include "common/x64/rdtsc.h"
+namespace Common::X64 {
+template <u64 Nearest>
+static u64 RoundToNearest(u64 value) {
+    const auto mod = value % Nearest;
+    return mod >= (Nearest / 2) ? (value - mod + Nearest) : (value - mod);
+}
+u64 EstimateRDTSCFrequency() {
+    // Discard the first result measuring the rdtsc.
+    FencedRDTSC();
+    std::this_thread::sleep_for(std::chrono::milliseconds{1});
+    FencedRDTSC();
+    // Get the current time.
+    const auto start_time = RealTimeClock::Now();
+    const u64 tsc_start = FencedRDTSC();
+    // Wait for 100 milliseconds.
+    std::this_thread::sleep_for(std::chrono::milliseconds{100});
+    const auto end_time = RealTimeClock::Now();
+    const u64 tsc_end = FencedRDTSC();
+    // Calculate differences.
+    const u64 timer_diff = static_cast<u64>(
+        std::chrono::duration_cast<std::chrono::nanoseconds>(end_time - start_time).count());
+    const u64 tsc_diff = tsc_end - tsc_start;
+    const u64 tsc_freq = MultiplyAndDivide64(tsc_diff, 1000000000ULL, timer_diff);
+    return RoundToNearest<100'000>(tsc_freq);
+}
+} // namespace Common::X64
diff --git a/src/common/x64/rdtsc.h b/src/common/x64/rdtsc.h
new file mode 100644
index 000000000..0ec4f52f9
--- /dev/null
+++ b/src/common/x64/rdtsc.h
@@ -0,0 +1,37 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+#pragma once
+#ifdef _MSC_VER
+#include <intrin.h>
+#endif
+#include "common/common_types.h"
+namespace Common::X64 {
+#ifdef _MSC_VER
+__forceinline static u64 FencedRDTSC() {
+    _mm_lfence();
+    _ReadWriteBarrier();
+    const u64 result = __rdtsc();
+    _mm_lfence();
+    _ReadWriteBarrier();
+    return result;
+}
+#else
+static inline u64 FencedRDTSC() {
+    u64 eax;
+    u64 edx;
+    asm volatile("lfence\n\t"
+                 "rdtsc\n\t"
+                 "lfence\n\t"
+                 : "=a"(eax), "=d"(edx));
+    return (edx << 32) | eax;
+}
+#endif
+u64 EstimateRDTSCFrequency();
+} // namespace Common::X64
diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt
index 227c431bc..3655b8478 100644
--- a/src/core/CMakeLists.txt
+++ b/src/core/CMakeLists.txt
@@ -14,7 +14,6 @@ add_library(core STATIC
    core.h
    core_timing.cpp
    core_timing.h
-    core_timing_util.h
    cpu_manager.cpp
    cpu_manager.h
    crypto/aes_util.cpp
diff --git a/src/core/core_timing.cpp b/src/core/core_timing.cpp
index 4f2692b05..4f0a3f8ea 100644
--- a/src/core/core_timing.cpp
+++ b/src/core/core_timing.cpp
@@ -16,12 +16,11 @@
 #include "common/microprofile.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 #include "core/hardware_properties.h"
 namespace Core::Timing {
-constexpr s64 MAX_SLICE_LENGTH = 4000;
+constexpr s64 MAX_SLICE_LENGTH = 10000;
 std::shared_ptr<EventType> CreateEvent(std::string name, TimedCallback&& callback) {
    return std::make_shared<EventType>(std::move(callback), std::move(name));
@@ -45,9 +44,7 @@ struct CoreTiming::Event {
    }
 };
-CoreTiming::CoreTiming()
+CoreTiming::CoreTiming() : clock{Common::CreateOptimalClock()} {}
-    : cpu_clock{Common::CreateBestMatchingClock(Hardware::BASE_CLOCK_RATE, Hardware::CNTFREQ)},
-      event_clock{Common::CreateStandardWallClock(Hardware::BASE_CLOCK_RATE, Hardware::CNTFREQ)} {}
 CoreTiming::~CoreTiming() {
    Reset();
@@ -68,7 +65,7 @@ void CoreTiming::Initialize(std::function<void()>&& on_thread_init_) {
    on_thread_init = std::move(on_thread_init_);
    event_fifo_id = 0;
    shutting_down = false;
-    ticks = 0;
+    cpu_ticks = 0;
    const auto empty_timed_callback = [](std::uintptr_t, u64, std::chrono::nanoseconds)
        -> std::optional<std::chrono::nanoseconds> { return std::nullopt; };
    ev_lost = CreateEvent("_lost_event", empty_timed_callback);
@@ -173,38 +170,30 @@ void CoreTiming::UnscheduleEvent(const std::shared_ptr<EventType>& event_type,
 }
 void CoreTiming::AddTicks(u64 ticks_to_add) {
-    ticks += ticks_to_add;
+    cpu_ticks += ticks_to_add;
-    downcount -= static_cast<s64>(ticks);
+    downcount -= static_cast<s64>(cpu_ticks);
 }
 void CoreTiming::Idle() {
-    if (!event_queue.empty()) {
+    cpu_ticks += 1000U;
-        const u64 next_event_time = event_queue.front().time;
-        const u64 next_ticks = nsToCycles(std::chrono::nanoseconds(next_event_time)) + 10U;
-        if (next_ticks > ticks) {
-            ticks = next_ticks;
-        }
-        return;
-    }
-    ticks += 1000U;
 }
 void CoreTiming::ResetTicks() {
    downcount = MAX_SLICE_LENGTH;
 }
-u64 CoreTiming::GetCPUTicks() const {
+u64 CoreTiming::GetClockTicks() const {
    if (is_multicore) [[likely]] {
-        return cpu_clock->GetCPUCycles();
+        return clock->GetCNTPCT();
    }
-    return ticks;
+    return Common::WallClock::CPUTickToCNTPCT(cpu_ticks);
 }
-u64 CoreTiming::GetClockTicks() const {
+u64 CoreTiming::GetGPUTicks() const {
    if (is_multicore) [[likely]] {
-        return cpu_clock->GetClockCycles();
+        return clock->GetGPUTick();
    }
-    return CpuCyclesToClockCycles(ticks);
+    return Common::WallClock::CPUTickToGPUTick(cpu_ticks);
 }
 std::optional<s64> CoreTiming::Advance() {
@@ -297,9 +286,7 @@ void CoreTiming::ThreadLoop() {
        }
        paused_set = true;
-        event_clock->Pause(true);
        pause_event.Wait();
-        event_clock->Pause(false);
    }
 }
@@ -315,25 +302,18 @@ void CoreTiming::Reset() {
    has_started = false;
 }
-std::chrono::nanoseconds CoreTiming::GetCPUTimeNs() const {
-    if (is_multicore) [[likely]] {
-        return cpu_clock->GetTimeNS();
-    }
-    return CyclesToNs(ticks);
-}
 std::chrono::nanoseconds CoreTiming::GetGlobalTimeNs() const {
    if (is_multicore) [[likely]] {
-        return event_clock->GetTimeNS();
+        return clock->GetTimeNS();
    }
-    return CyclesToNs(ticks);
+    return std::chrono::nanoseconds{Common::WallClock::CPUTickToNS(cpu_ticks)};
 }
 std::chrono::microseconds CoreTiming::GetGlobalTimeUs() const {
    if (is_multicore) [[likely]] {
-        return event_clock->GetTimeUS();
+        return clock->GetTimeUS();
    }
-    return CyclesToUs(ticks);
+    return std::chrono::microseconds{Common::WallClock::CPUTickToUS(cpu_ticks)};
 }
 } // namespace Core::Timing
diff --git a/src/core/core_timing.h b/src/core/core_timing.h
index e7c4a949f..10db1de55 100644
--- a/src/core/core_timing.h
+++ b/src/core/core_timing.h
@@ -116,14 +116,11 @@ public:
        return downcount;
    }
-    /// Returns current time in emulated CPU cycles
+    /// Returns the current CNTPCT tick value.
-    u64 GetCPUTicks() const;
-    /// Returns current time in emulated in Clock cycles
    u64 GetClockTicks() const;
-    /// Returns current time in nanoseconds.
+    /// Returns the current GPU tick value.
-    std::chrono::nanoseconds GetCPUTimeNs() const;
+    u64 GetGPUTicks() const;
    /// Returns current time in microseconds.
    std::chrono::microseconds GetGlobalTimeUs() const;
@@ -142,8 +139,7 @@ private:
    void Reset();
-    std::unique_ptr<Common::WallClock> cpu_clock;
+    std::unique_ptr<Common::WallClock> clock;
-    std::unique_ptr<Common::WallClock> event_clock;
    s64 global_timer = 0;
@@ -171,7 +167,7 @@ private:
    s64 pause_end_time{};
    /// Cycle timing
-    u64 ticks{};
+    u64 cpu_ticks{};
    s64 downcount{};
 };
diff --git a/src/core/core_timing_util.h b/src/core/core_timing_util.h
deleted file mode 100644
index fe5aaefc7..000000000
--- a/src/core/core_timing_util.h
+++ /dev/null
@@ -1,58 +0,0 @@
-// SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
-// SPDX-License-Identifier: GPL-2.0-or-later
-#pragma once
-#include <chrono>
-#include "common/common_types.h"
-#include "core/hardware_properties.h"
-namespace Core::Timing {
-namespace detail {
-constexpr u64 CNTFREQ_ADJUSTED = Hardware::CNTFREQ / 1000;
-constexpr u64 BASE_CLOCK_RATE_ADJUSTED = Hardware::BASE_CLOCK_RATE / 1000;
-} // namespace detail
-[[nodiscard]] constexpr s64 msToCycles(std::chrono::milliseconds ms) {
-    return ms.count() * detail::BASE_CLOCK_RATE_ADJUSTED;
-}
-[[nodiscard]] constexpr s64 usToCycles(std::chrono::microseconds us) {
-    return us.count() * detail::BASE_CLOCK_RATE_ADJUSTED / 1000;
-}
-[[nodiscard]] constexpr s64 nsToCycles(std::chrono::nanoseconds ns) {
-    return ns.count() * detail::BASE_CLOCK_RATE_ADJUSTED / 1000000;
-}
-[[nodiscard]] constexpr u64 msToClockCycles(std::chrono::milliseconds ms) {
-    return static_cast<u64>(ms.count()) * detail::CNTFREQ_ADJUSTED;
-}
-[[nodiscard]] constexpr u64 usToClockCycles(std::chrono::microseconds us) {
-    return us.count() * detail::CNTFREQ_ADJUSTED / 1000;
-}
-[[nodiscard]] constexpr u64 nsToClockCycles(std::chrono::nanoseconds ns) {
-    return ns.count() * detail::CNTFREQ_ADJUSTED / 1000000;
-}
-[[nodiscard]] constexpr u64 CpuCyclesToClockCycles(u64 ticks) {
-    return ticks * detail::CNTFREQ_ADJUSTED / detail::BASE_CLOCK_RATE_ADJUSTED;
-}
-[[nodiscard]] constexpr std::chrono::milliseconds CyclesToMs(s64 cycles) {
-    return std::chrono::milliseconds(cycles / detail::BASE_CLOCK_RATE_ADJUSTED);
-}
-[[nodiscard]] constexpr std::chrono::nanoseconds CyclesToNs(s64 cycles) {
-    return std::chrono::nanoseconds(cycles * 1000000 / detail::BASE_CLOCK_RATE_ADJUSTED);
-}
-[[nodiscard]] constexpr std::chrono::microseconds CyclesToUs(s64 cycles) {
-    return std::chrono::microseconds(cycles * 1000 / detail::BASE_CLOCK_RATE_ADJUSTED);
-}
-} // namespace Core::Timing
diff --git a/src/core/file_sys/patch_manager.cpp b/src/core/file_sys/patch_manager.cpp
index 4e61d4335..d3286b352 100644
--- a/src/core/file_sys/patch_manager.cpp
+++ b/src/core/file_sys/patch_manager.cpp
@@ -153,7 +153,7 @@ VirtualDir PatchManager::PatchExeFS(VirtualDir exefs) const {
    const auto sdmc_load_dir = fs_controller.GetSDMCModificationLoadRoot(title_id);
    std::vector<VirtualDir> patch_dirs = {sdmc_load_dir};
-    if (load_dir != nullptr && load_dir->GetSize() > 0) {
+    if (load_dir != nullptr) {
        const auto load_patch_dirs = load_dir->GetSubdirectories();
        patch_dirs.insert(patch_dirs.end(), load_patch_dirs.begin(), load_patch_dirs.end());
    }
@@ -354,8 +354,7 @@ static void ApplyLayeredFS(VirtualFile& romfs, u64 title_id, ContentRecordType t
    const auto load_dir = fs_controller.GetModificationLoadRoot(title_id);
    const auto sdmc_load_dir = fs_controller.GetSDMCModificationLoadRoot(title_id);
    if ((type != ContentRecordType::Program && type != ContentRecordType::Data) ||
-        ((load_dir == nullptr || load_dir->GetSize() <= 0) &&
+        (load_dir == nullptr && sdmc_load_dir == nullptr)) {
-         (sdmc_load_dir == nullptr || sdmc_load_dir->GetSize() <= 0))) {
        return;
    }
@@ -496,7 +495,7 @@ PatchManager::PatchVersionNames PatchManager::GetPatchVersionNames(VirtualFile u
    // General Mods (LayeredFS and IPS)
    const auto mod_dir = fs_controller.GetModificationLoadRoot(title_id);
-    if (mod_dir != nullptr && mod_dir->GetSize() > 0) {
+    if (mod_dir != nullptr) {
        for (const auto& mod : mod_dir->GetSubdirectories()) {
            std::string types;
@@ -540,7 +539,7 @@ PatchManager::PatchVersionNames PatchManager::GetPatchVersionNames(VirtualFile u
    // SDMC mod directory (RomFS LayeredFS)
    const auto sdmc_mod_dir = fs_controller.GetSDMCModificationLoadRoot(title_id);
-    if (sdmc_mod_dir != nullptr && sdmc_mod_dir->GetSize() > 0) {
+    if (sdmc_mod_dir != nullptr) {
        std::string types;
        if (IsDirValidAndNonEmpty(FindSubdirectoryCaseless(sdmc_mod_dir, "exefs"))) {
            AppendCommaIfNotEmpty(types, "LayeredExeFS");
diff --git a/src/core/file_sys/vfs_concat.cpp b/src/core/file_sys/vfs_concat.cpp
index 853b893a1..311a59e5f 100644
--- a/src/core/file_sys/vfs_concat.cpp
+++ b/src/core/file_sys/vfs_concat.cpp
@@ -150,23 +150,29 @@ std::size_t ConcatenatedVfsFile::Read(u8* data, std::size_t length, std::size_t
    while (cur_length > 0 && it != concatenation_map.end()) {
        // Check if we can read the file at this position.
        const auto& file = it->file;
-        const u64 file_offset = it->offset;
+        const u64 map_offset = it->offset;
        const u64 file_size = file->GetSize();
-        if (cur_offset >= file_offset + file_size) {
+        if (cur_offset > map_offset + file_size) {
            // Entirely out of bounds read.
            break;
        }
        // Read the file at this position.
-        const u64 intended_read_size = std::min<u64>(cur_length, file_size);
+        const u64 file_seek = cur_offset - map_offset;
+        const u64 intended_read_size = std::min<u64>(cur_length, file_size - file_seek);
        const u64 actual_read_size =
-            file->Read(data + (cur_offset - offset), intended_read_size, cur_offset - file_offset);
+            file->Read(data + (cur_offset - offset), intended_read_size, file_seek);
        // Update tracking.
        cur_offset += actual_read_size;
        cur_length -= actual_read_size;
        it++;
+        // If we encountered a short read, we're done.
+        if (actual_read_size < intended_read_size) {
+            break;
+        }
    }
    return cur_offset - offset;
diff --git a/src/core/file_sys/vfs_real.cpp b/src/core/file_sys/vfs_real.cpp
index 7a15d8438..b0515ec05 100644
--- a/src/core/file_sys/vfs_real.cpp
+++ b/src/core/file_sys/vfs_real.cpp
@@ -10,6 +10,7 @@
 #include "common/fs/fs.h"
 #include "common/fs/path_util.h"
 #include "common/logging/log.h"
+#include "core/file_sys/vfs.h"
 #include "core/file_sys/vfs_real.h"
 // For FileTimeStampRaw
@@ -72,8 +73,10 @@ VfsEntryType RealVfsFilesystem::GetEntryType(std::string_view path_) const {
    return VfsEntryType::File;
 }
-VirtualFile RealVfsFilesystem::OpenFile(std::string_view path_, Mode perms) {
+VirtualFile RealVfsFilesystem::OpenFileFromEntry(std::string_view path_, std::optional<u64> size,
+                                                 Mode perms) {
    const auto path = FS::SanitizePath(path_, FS::DirectorySeparator::PlatformDefault);
+    std::scoped_lock lk{list_lock};
    if (auto it = cache.find(path); it != cache.end()) {
        if (auto file = it->second.lock(); file) {
@@ -81,23 +84,30 @@ VirtualFile RealVfsFilesystem::OpenFile(std::string_view path_, Mode perms) {
        }
    }
-    if (!FS::Exists(path) || !FS::IsFile(path)) {
+    if (!size && !FS::IsFile(path)) {
        return nullptr;
    }
    auto reference = std::make_unique<FileReference>();
-    this->InsertReferenceIntoList(*reference);
+    this->InsertReferenceIntoListLocked(*reference);
-    auto file =
+    auto file = std::shared_ptr<RealVfsFile>(
-        std::shared_ptr<RealVfsFile>(new RealVfsFile(*this, std::move(reference), path, perms));
+        new RealVfsFile(*this, std::move(reference), path, perms, size));
    cache[path] = file;
    return file;
 }
+VirtualFile RealVfsFilesystem::OpenFile(std::string_view path_, Mode perms) {
+    return OpenFileFromEntry(path_, {}, perms);
+}
 VirtualFile RealVfsFilesystem::CreateFile(std::string_view path_, Mode perms) {
    const auto path = FS::SanitizePath(path_, FS::DirectorySeparator::PlatformDefault);
-    cache.erase(path);
+    {
+        std::scoped_lock lk{list_lock};
+        cache.erase(path);
+    }
    // Current usages of CreateFile expect to delete the contents of an existing file.
    if (FS::IsFile(path)) {
@@ -127,8 +137,11 @@ VirtualFile RealVfsFilesystem::CopyFile(std::string_view old_path_, std::string_
 VirtualFile RealVfsFilesystem::MoveFile(std::string_view old_path_, std::string_view new_path_) {
    const auto old_path = FS::SanitizePath(old_path_, FS::DirectorySeparator::PlatformDefault);
    const auto new_path = FS::SanitizePath(new_path_, FS::DirectorySeparator::PlatformDefault);
-    cache.erase(old_path);
+    {
-    cache.erase(new_path);
+        std::scoped_lock lk{list_lock};
+        cache.erase(old_path);
+        cache.erase(new_path);
+    }
    if (!FS::RenameFile(old_path, new_path)) {
        return nullptr;
    }
@@ -137,7 +150,10 @@ VirtualFile RealVfsFilesystem::MoveFile(std::string_view old_path_, std::string_
 bool RealVfsFilesystem::DeleteFile(std::string_view path_) {
    const auto path = FS::SanitizePath(path_, FS::DirectorySeparator::PlatformDefault);
-    cache.erase(path);
+    {
+        std::scoped_lock lk{list_lock};
+        cache.erase(path);
+    }
    return FS::RemoveFile(path);
 }
@@ -176,14 +192,17 @@ bool RealVfsFilesystem::DeleteDirectory(std::string_view path_) {
    return FS::RemoveDirRecursively(path);
 }
-void RealVfsFilesystem::RefreshReference(const std::string& path, Mode perms,
+std::unique_lock<std::mutex> RealVfsFilesystem::RefreshReference(const std::string& path,
-                                         FileReference& reference) {
+                                                                 Mode perms,
+                                                                 FileReference& reference) {
+    std::unique_lock lk{list_lock};
    // Temporarily remove from list.
-    this->RemoveReferenceFromList(reference);
+    this->RemoveReferenceFromListLocked(reference);
    // Restore file if needed.
    if (!reference.file) {
-        this->EvictSingleReference();
+        this->EvictSingleReferenceLocked();
        reference.file =
            FS::FileOpen(path, ModeFlagsToFileAccessMode(perms), FS::FileType::BinaryFile);
@@ -193,12 +212,16 @@ void RealVfsFilesystem::RefreshReference(const std::string& path, Mode perms,
    }
    // Reinsert into list.
-    this->InsertReferenceIntoList(reference);
+    this->InsertReferenceIntoListLocked(reference);
+    return lk;
 }
 void RealVfsFilesystem::DropReference(std::unique_ptr<FileReference>&& reference) {
+    std::scoped_lock lk{list_lock};
    // Remove from list.
-    this->RemoveReferenceFromList(*reference);
+    this->RemoveReferenceFromListLocked(*reference);
    // Close the file.
    if (reference->file) {
@@ -207,14 +230,14 @@ void RealVfsFilesystem::DropReference(std::unique_ptr<FileReference>&& reference
    }
 }
-void RealVfsFilesystem::EvictSingleReference() {
+void RealVfsFilesystem::EvictSingleReferenceLocked() {
    if (num_open_files < MaxOpenFiles || open_references.empty()) {
        return;
    }
    // Get and remove from list.
    auto& reference = open_references.back();
-    this->RemoveReferenceFromList(reference);
+    this->RemoveReferenceFromListLocked(reference);
    // Close the file.
    if (reference.file) {
@@ -223,10 +246,10 @@ void RealVfsFilesystem::EvictSingleReference() {
    }
    // Reinsert into closed list.
-    this->InsertReferenceIntoList(reference);
+    this->InsertReferenceIntoListLocked(reference);
 }
-void RealVfsFilesystem::InsertReferenceIntoList(FileReference& reference) {
+void RealVfsFilesystem::InsertReferenceIntoListLocked(FileReference& reference) {
    if (reference.file) {
        open_references.push_front(reference);
    } else {
@@ -234,7 +257,7 @@ void RealVfsFilesystem::InsertReferenceIntoList(FileReference& reference) {
    }
 }
-void RealVfsFilesystem::RemoveReferenceFromList(FileReference& reference) {
+void RealVfsFilesystem::RemoveReferenceFromListLocked(FileReference& reference) {
    if (reference.file) {
        open_references.erase(open_references.iterator_to(reference));
    } else {
@@ -243,10 +266,10 @@ void RealVfsFilesystem::RemoveReferenceFromList(FileReference& reference) {
 }
 RealVfsFile::RealVfsFile(RealVfsFilesystem& base_, std::unique_ptr<FileReference> reference_,
-                         const std::string& path_, Mode perms_)
+                         const std::string& path_, Mode perms_, std::optional<u64> size_)
    : base(base_), reference(std::move(reference_)), path(path_),
      parent_path(FS::GetParentPath(path_)), path_components(FS::SplitPathComponents(path_)),
-      perms(perms_) {}
+      size(size_), perms(perms_) {}
 RealVfsFile::~RealVfsFile() {
    base.DropReference(std::move(reference));
@@ -257,12 +280,15 @@ std::string RealVfsFile::GetName() const {
 }
 std::size_t RealVfsFile::GetSize() const {
-    base.RefreshReference(path, perms, *reference);
+    if (size) {
-    return reference->file ? reference->file->GetSize() : 0;
+        return *size;
+    }
+    return FS::GetSize(path);
 }
 bool RealVfsFile::Resize(std::size_t new_size) {
-    base.RefreshReference(path, perms, *reference);
+    size.reset();
+    auto lk = base.RefreshReference(path, perms, *reference);
    return reference->file ? reference->file->SetSize(new_size) : false;
 }
@@ -279,7 +305,7 @@ bool RealVfsFile::IsReadable() const {
 }
 std::size_t RealVfsFile::Read(u8* data, std::size_t length, std::size_t offset) const {
-    base.RefreshReference(path, perms, *reference);
+    auto lk = base.RefreshReference(path, perms, *reference);
    if (!reference->file || !reference->file->Seek(static_cast<s64>(offset))) {
        return 0;
    }
@@ -287,7 +313,8 @@ std::size_t RealVfsFile::Read(u8* data, std::size_t length, std::size_t offset)
 }
 std::size_t RealVfsFile::Write(const u8* data, std::size_t length, std::size_t offset) {
-    base.RefreshReference(path, perms, *reference);
+    size.reset();
+    auto lk = base.RefreshReference(path, perms, *reference);
    if (!reference->file || !reference->file->Seek(static_cast<s64>(offset))) {
        return 0;
    }
@@ -309,10 +336,11 @@ std::vector<VirtualFile> RealVfsDirectory::IterateEntries<RealVfsFile, VfsFile>(
    std::vector<VirtualFile> out;
-    const FS::DirEntryCallable callback = [this, &out](const std::filesystem::path& full_path) {
+    const FS::DirEntryCallable callback = [this,
-        const auto full_path_string = FS::PathToUTF8String(full_path);
+                                           &out](const std::filesystem::directory_entry& entry) {
+        const auto full_path_string = FS::PathToUTF8String(entry.path());
-        out.emplace_back(base.OpenFile(full_path_string, perms));
+        out.emplace_back(base.OpenFileFromEntry(full_path_string, entry.file_size(), perms));
        return true;
    };
@@ -330,8 +358,9 @@ std::vector<VirtualDir> RealVfsDirectory::IterateEntries<RealVfsDirectory, VfsDi
    std::vector<VirtualDir> out;
-    const FS::DirEntryCallable callback = [this, &out](const std::filesystem::path& full_path) {
+    const FS::DirEntryCallable callback = [this,
-        const auto full_path_string = FS::PathToUTF8String(full_path);
+                                           &out](const std::filesystem::directory_entry& entry) {
+        const auto full_path_string = FS::PathToUTF8String(entry.path());
        out.emplace_back(base.OpenDirectory(full_path_string, perms));
@@ -483,12 +512,10 @@ std::map<std::string, VfsEntryType, std::less<>> RealVfsDirectory::GetEntries()
    std::map<std::string, VfsEntryType, std::less<>> out;
-    const FS::DirEntryCallable callback = [&out](const std::filesystem::path& full_path) {
+    const FS::DirEntryCallable callback = [&out](const std::filesystem::directory_entry& entry) {
-        const auto filename = FS::PathToUTF8String(full_path.filename());
+        const auto filename = FS::PathToUTF8String(entry.path().filename());
        out.insert_or_assign(filename,
-                             FS::IsDir(full_path) ? VfsEntryType::Directory : VfsEntryType::File);
+                             entry.is_directory() ? VfsEntryType::Directory : VfsEntryType::File);
        return true;
    };
diff --git a/src/core/file_sys/vfs_real.h b/src/core/file_sys/vfs_real.h
index d8c900e33..26ea7df62 100644
--- a/src/core/file_sys/vfs_real.h
+++ b/src/core/file_sys/vfs_real.h
@@ -4,6 +4,8 @@
 #pragma once
 #include <map>
+#include <mutex>
+#include <optional>
 #include <string_view>
 #include "common/intrusive_list.h"
 #include "core/file_sys/mode.h"
@@ -20,6 +22,8 @@ struct FileReference : public Common::IntrusiveListBaseNode<FileReference> {
 };
 class RealVfsFile;
+class RealVfsDirectory;
 class RealVfsFilesystem : public VfsFilesystem {
 public:
    RealVfsFilesystem();
@@ -45,17 +49,24 @@ private:
    std::map<std::string, std::weak_ptr<VfsFile>, std::less<>> cache;
    ReferenceListType open_references;
    ReferenceListType closed_references;
+    std::mutex list_lock;
    size_t num_open_files{};
 private:
    friend class RealVfsFile;
-    void RefreshReference(const std::string& path, Mode perms, FileReference& reference);
+    std::unique_lock<std::mutex> RefreshReference(const std::string& path, Mode perms,
+                                                  FileReference& reference);
    void DropReference(std::unique_ptr<FileReference>&& reference);
-    void EvictSingleReference();
 private:
-    void InsertReferenceIntoList(FileReference& reference);
+    friend class RealVfsDirectory;
-    void RemoveReferenceFromList(FileReference& reference);
+    VirtualFile OpenFileFromEntry(std::string_view path, std::optional<u64> size,
+                                  Mode perms = Mode::Read);
+private:
+    void EvictSingleReferenceLocked();
+    void InsertReferenceIntoListLocked(FileReference& reference);
+    void RemoveReferenceFromListLocked(FileReference& reference);
 };
 // An implementation of VfsFile that represents a file on the user's computer.
@@ -78,13 +89,14 @@ public:
 private:
    RealVfsFile(RealVfsFilesystem& base, std::unique_ptr<FileReference> reference,
-                const std::string& path, Mode perms = Mode::Read);
+                const std::string& path, Mode perms = Mode::Read, std::optional<u64> size = {});
    RealVfsFilesystem& base;
    std::unique_ptr<FileReference> reference;
    std::string path;
    std::string parent_path;
    std::vector<std::string> path_components;
+    std::optional<u64> size;
    Mode perms;
 };
diff --git a/src/core/hid/emulated_controller.cpp b/src/core/hid/emulated_controller.cpp
index 0a7777732..c937495f9 100644
--- a/src/core/hid/emulated_controller.cpp
+++ b/src/core/hid/emulated_controller.cpp
@@ -149,12 +149,16 @@ void EmulatedController::LoadDevices() {
    camera_params[0] = right_joycon;
    camera_params[0].Set("camera", true);
-    camera_params[1] = Common::ParamPackage{"engine:camera,camera:1"};
-    ring_params[1] = Common::ParamPackage{"engine:joycon,axis_x:100,axis_y:101"};
-    nfc_params[0] = Common::ParamPackage{"engine:virtual_amiibo,nfc:1"};
    nfc_params[1] = right_joycon;
    nfc_params[1].Set("nfc", true);
+    // Only map virtual devices to the first controller
+    if (npad_id_type == NpadIdType::Player1 || npad_id_type == NpadIdType::Handheld) {
+        camera_params[1] = Common::ParamPackage{"engine:camera,camera:1"};
+        ring_params[1] = Common::ParamPackage{"engine:joycon,axis_x:100,axis_y:101"};
+        nfc_params[0] = Common::ParamPackage{"engine:virtual_amiibo,nfc:1"};
+    }
    output_params[LeftIndex] = left_joycon;
    output_params[RightIndex] = right_joycon;
    output_params[2] = camera_params[1];
@@ -1176,10 +1180,7 @@ void EmulatedController::SetNfc(const Common::Input::CallbackStatus& callback) {
        return;
    }
-    controller.nfc_state = {
+    controller.nfc_state = controller.nfc_values;
-        controller.nfc_values.state,
-        controller.nfc_values.data,
-    };
 }
 bool EmulatedController::SetVibration(std::size_t device_index, VibrationValue vibration) {
@@ -1308,6 +1309,73 @@ bool EmulatedController::HasNfc() const {
    return is_connected && (has_virtual_nfc && is_virtual_nfc_supported);
 }
+bool EmulatedController::AddNfcHandle() {
+    nfc_handles++;
+    return SetPollingMode(EmulatedDeviceIndex::RightIndex, Common::Input::PollingMode::NFC) ==
+           Common::Input::DriverResult::Success;
+}
+bool EmulatedController::RemoveNfcHandle() {
+    nfc_handles--;
+    if (nfc_handles <= 0) {
+        return SetPollingMode(EmulatedDeviceIndex::RightIndex,
+                              Common::Input::PollingMode::Active) ==
+               Common::Input::DriverResult::Success;
+    }
+    return true;
+}
+bool EmulatedController::StartNfcPolling() {
+    auto& nfc_output_device = output_devices[static_cast<std::size_t>(DeviceIndex::Right)];
+    auto& nfc_virtual_output_device = output_devices[3];
+    return nfc_output_device->StartNfcPolling() == Common::Input::NfcState::Success ||
+           nfc_virtual_output_device->StartNfcPolling() == Common::Input::NfcState::Success;
+}
+bool EmulatedController::StopNfcPolling() {
+    auto& nfc_output_device = output_devices[static_cast<std::size_t>(DeviceIndex::Right)];
+    auto& nfc_virtual_output_device = output_devices[3];
+    return nfc_output_device->StopNfcPolling() == Common::Input::NfcState::Success ||
+           nfc_virtual_output_device->StopNfcPolling() == Common::Input::NfcState::Success;
+}
+bool EmulatedController::ReadAmiiboData(std::vector<u8>& data) {
+    auto& nfc_output_device = output_devices[static_cast<std::size_t>(DeviceIndex::Right)];
+    auto& nfc_virtual_output_device = output_devices[3];
+    if (nfc_output_device->ReadAmiiboData(data) == Common::Input::NfcState::Success) {
+        return true;
+    }
+    return nfc_virtual_output_device->ReadAmiiboData(data) == Common::Input::NfcState::Success;
+}
+bool EmulatedController::ReadMifareData(const Common::Input::MifareRequest& request,
+                                        Common::Input::MifareRequest& out_data) {
+    auto& nfc_output_device = output_devices[static_cast<std::size_t>(DeviceIndex::Right)];
+    auto& nfc_virtual_output_device = output_devices[3];
+    if (nfc_output_device->ReadMifareData(request, out_data) == Common::Input::NfcState::Success) {
+        return true;
+    }
+    return nfc_virtual_output_device->ReadMifareData(request, out_data) ==
+           Common::Input::NfcState::Success;
+}
+bool EmulatedController::WriteMifareData(const Common::Input::MifareRequest& request) {
+    auto& nfc_output_device = output_devices[static_cast<std::size_t>(DeviceIndex::Right)];
+    auto& nfc_virtual_output_device = output_devices[3];
+    if (nfc_output_device->WriteMifareData(request) == Common::Input::NfcState::Success) {
+        return true;
+    }
+    return nfc_virtual_output_device->WriteMifareData(request) == Common::Input::NfcState::Success;
+}
 bool EmulatedController::WriteNfc(const std::vector<u8>& data) {
    auto& nfc_output_device = output_devices[static_cast<std::size_t>(DeviceIndex::Right)];
    auto& nfc_virtual_output_device = output_devices[3];
diff --git a/src/core/hid/emulated_controller.h b/src/core/hid/emulated_controller.h
index 09fe1a0ab..d511e5fac 100644
--- a/src/core/hid/emulated_controller.h
+++ b/src/core/hid/emulated_controller.h
@@ -97,10 +97,7 @@ struct RingSensorForce {
    f32 force;
 };
-struct NfcState {
+using NfcState = Common::Input::NfcStatus;
-    Common::Input::NfcState state{};
-    std::vector<u8> data{};
-};
 struct ControllerMotion {
    Common::Vec3f accel{};
@@ -393,9 +390,31 @@ public:
    /// Returns true if the device has nfc support
    bool HasNfc() const;
+    /// Sets the joycon in nfc mode and increments the handle count
+    bool AddNfcHandle();
+    /// Decrements the handle count if zero sets the joycon in active mode
+    bool RemoveNfcHandle();
+    /// Start searching for nfc tags
+    bool StartNfcPolling();
+    /// Stop searching for nfc tags
+    bool StopNfcPolling();
+    /// Returns true if the nfc tag was readable
+    bool ReadAmiiboData(std::vector<u8>& data);
    /// Returns true if the nfc tag was written
    bool WriteNfc(const std::vector<u8>& data);
+    /// Returns true if the nfc tag was readable
+    bool ReadMifareData(const Common::Input::MifareRequest& request,
+                        Common::Input::MifareRequest& out_data);
+    /// Returns true if the nfc tag was written
+    bool WriteMifareData(const Common::Input::MifareRequest& request);
    /// Returns the led pattern corresponding to this emulated controller
    LedPattern GetLedPattern() const;
@@ -532,6 +551,7 @@ private:
    bool system_buttons_enabled{true};
    f32 motion_sensitivity{Core::HID::MotionInput::IsAtRestStandard};
    u32 turbo_button_state{0};
+    std::size_t nfc_handles{0};
    // Temporary values to avoid doing changes while the controller is in configuring mode
    NpadStyleIndex tmp_npad_type{NpadStyleIndex::None};
diff --git a/src/core/hid/input_converter.cpp b/src/core/hid/input_converter.cpp
index 4ccb1c596..a05716fd8 100644
--- a/src/core/hid/input_converter.cpp
+++ b/src/core/hid/input_converter.cpp
@@ -299,11 +299,7 @@ Common::Input::NfcStatus TransformToNfc(const Common::Input::CallbackStatus& cal
    Common::Input::NfcStatus nfc{};
    switch (callback.type) {
    case Common::Input::InputType::Nfc:
-        nfc = {
+        return callback.nfc_status;
-            .state = callback.nfc_status,
-            .data = callback.raw_data,
-        };
-        break;
    default:
        LOG_ERROR(Input, "Conversion from type {} to NFC not implemented", callback.type);
        break;
diff --git a/src/core/hle/kernel/k_scheduler.cpp b/src/core/hle/kernel/k_scheduler.cpp
index faa12b4f0..75ce5a23c 100644
--- a/src/core/hle/kernel/k_scheduler.cpp
+++ b/src/core/hle/kernel/k_scheduler.cpp
@@ -184,7 +184,8 @@ u64 KScheduler::UpdateHighestPriorityThread(KThread* highest_thread) {
        prev_highest_thread != highest_thread) [[likely]] {
        if (prev_highest_thread != nullptr) [[likely]] {
            IncrementScheduledCount(prev_highest_thread);
-            prev_highest_thread->SetLastScheduledTick(m_kernel.System().CoreTiming().GetCPUTicks());
+            prev_highest_thread->SetLastScheduledTick(
+                m_kernel.System().CoreTiming().GetClockTicks());
        }
        if (m_state.should_count_idle) {
            if (highest_thread != nullptr) [[likely]] {
@@ -351,7 +352,7 @@ void KScheduler::SwitchThread(KThread* next_thread) {
    // Update the CPU time tracking variables.
    const s64 prev_tick = m_last_context_switch_time;
-    const s64 cur_tick = m_kernel.System().CoreTiming().GetCPUTicks();
+    const s64 cur_tick = m_kernel.System().CoreTiming().GetClockTicks();
    const s64 tick_diff = cur_tick - prev_tick;
    cur_thread->AddCpuTime(m_core_id, tick_diff);
    if (cur_process != nullptr) {
diff --git a/src/core/hle/kernel/k_synchronization_object.cpp b/src/core/hle/kernel/k_synchronization_object.cpp
index b7da3eee7..3e5b735b1 100644
--- a/src/core/hle/kernel/k_synchronization_object.cpp
+++ b/src/core/hle/kernel/k_synchronization_object.cpp
@@ -3,6 +3,7 @@
 #include "common/assert.h"
 #include "common/common_types.h"
+#include "common/scratch_buffer.h"
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/k_scoped_scheduler_lock_and_sleep.h"
 #include "core/hle/kernel/k_synchronization_object.h"
@@ -75,7 +76,7 @@ Result KSynchronizationObject::Wait(KernelCore& kernel, s32* out_index,
                                    KSynchronizationObject** objects, const s32 num_objects,
                                    s64 timeout) {
    // Allocate space on stack for thread nodes.
-    std::vector<ThreadListNode> thread_nodes(num_objects);
+    std::array<ThreadListNode, Svc::ArgumentHandleCountMax> thread_nodes;
    // Prepare for wait.
    KThread* thread = GetCurrentThreadPointer(kernel);
diff --git a/src/core/hle/kernel/k_thread.cpp b/src/core/hle/kernel/k_thread.cpp
index 908811e2c..adb6ec581 100644
--- a/src/core/hle/kernel/k_thread.cpp
+++ b/src/core/hle/kernel/k_thread.cpp
@@ -909,7 +909,7 @@ Result KThread::SetActivity(Svc::ThreadActivity activity) {
    R_SUCCEED();
 }
-Result KThread::GetThreadContext3(std::vector<u8>& out) {
+Result KThread::GetThreadContext3(Common::ScratchBuffer<u8>& out) {
    // Lock ourselves.
    KScopedLightLock lk{m_activity_pause_lock};
@@ -927,15 +927,13 @@ Result KThread::GetThreadContext3(std::vector<u8>& out) {
                // Mask away mode bits, interrupt bits, IL bit, and other reserved bits.
                auto context = GetContext64();
                context.pstate &= 0xFF0FFE20;
+                out.resize_destructive(sizeof(context));
-                out.resize(sizeof(context));
                std::memcpy(out.data(), std::addressof(context), sizeof(context));
            } else {
                // Mask away mode bits, interrupt bits, IL bit, and other reserved bits.
                auto context = GetContext32();
                context.cpsr &= 0xFF0FFE20;
+                out.resize_destructive(sizeof(context));
-                out.resize(sizeof(context));
                std::memcpy(out.data(), std::addressof(context), sizeof(context));
            }
        }
diff --git a/src/core/hle/kernel/k_thread.h b/src/core/hle/kernel/k_thread.h
index 37fe5db77..dd662b3f8 100644
--- a/src/core/hle/kernel/k_thread.h
+++ b/src/core/hle/kernel/k_thread.h
@@ -15,6 +15,7 @@
 #include "common/intrusive_list.h"
 #include "common/intrusive_red_black_tree.h"
+#include "common/scratch_buffer.h"
 #include "common/spin_lock.h"
 #include "core/arm/arm_interface.h"
 #include "core/hle/kernel/k_affinity_mask.h"
@@ -567,7 +568,7 @@ public:
    void RemoveWaiter(KThread* thread);
-    Result GetThreadContext3(std::vector<u8>& out);
+    Result GetThreadContext3(Common::ScratchBuffer<u8>& out);
    KThread* RemoveUserWaiterByKey(bool* out_has_waiters, KProcessAddress key) {
        return this->RemoveWaiterByKey(out_has_waiters, key, false);
diff --git a/src/core/hle/kernel/svc/svc_info.cpp b/src/core/hle/kernel/svc/svc_info.cpp
index 2b2c878b5..445cdd87b 100644
--- a/src/core/hle/kernel/svc/svc_info.cpp
+++ b/src/core/hle/kernel/svc/svc_info.cpp
@@ -199,9 +199,9 @@ Result GetInfo(Core::System& system, u64* result, InfoType info_id_type, Handle
        if (same_thread && info_sub_id == 0xFFFFFFFFFFFFFFFF) {
            const u64 thread_ticks = current_thread->GetCpuTime();
-            out_ticks = thread_ticks + (core_timing.GetCPUTicks() - prev_ctx_ticks);
+            out_ticks = thread_ticks + (core_timing.GetClockTicks() - prev_ctx_ticks);
        } else if (same_thread && info_sub_id == system.Kernel().CurrentPhysicalCoreIndex()) {
-            out_ticks = core_timing.GetCPUTicks() - prev_ctx_ticks;
+            out_ticks = core_timing.GetClockTicks() - prev_ctx_ticks;
        }
        *result = out_ticks;
diff --git a/src/core/hle/kernel/svc/svc_ipc.cpp b/src/core/hle/kernel/svc/svc_ipc.cpp
index ea03068aa..60247df2e 100644
--- a/src/core/hle/kernel/svc/svc_ipc.cpp
+++ b/src/core/hle/kernel/svc/svc_ipc.cpp
@@ -2,6 +2,7 @@
 // SPDX-License-Identifier: GPL-2.0-or-later
 #include "common/scope_exit.h"
+#include "common/scratch_buffer.h"
 #include "core/core.h"
 #include "core/hle/kernel/k_client_session.h"
 #include "core/hle/kernel/k_process.h"
@@ -45,11 +46,11 @@ Result ReplyAndReceive(Core::System& system, s32* out_index, uint64_t handles_ad
                 handles_addr, static_cast<u64>(sizeof(Handle) * num_handles)),
             ResultInvalidPointer);
-    std::vector<Handle> handles(num_handles);
+    std::array<Handle, Svc::ArgumentHandleCountMax> handles;
    GetCurrentMemory(kernel).ReadBlock(handles_addr, handles.data(), sizeof(Handle) * num_handles);
    // Convert handle list to object table.
-    std::vector<KSynchronizationObject*> objs(num_handles);
+    std::array<KSynchronizationObject*, Svc::ArgumentHandleCountMax> objs;
    R_UNLESS(handle_table.GetMultipleObjects<KSynchronizationObject>(objs.data(), handles.data(),
                                                                     num_handles),
             ResultInvalidHandle);
@@ -80,7 +81,7 @@ Result ReplyAndReceive(Core::System& system, s32* out_index, uint64_t handles_ad
        // Wait for an object.
        s32 index;
        Result result = KSynchronizationObject::Wait(kernel, std::addressof(index), objs.data(),
-                                                     static_cast<s32>(objs.size()), timeout_ns);
+                                                     num_handles, timeout_ns);
        if (result == ResultTimedOut) {
            R_RETURN(result);
        }
diff --git a/src/core/hle/kernel/svc/svc_synchronization.cpp b/src/core/hle/kernel/svc/svc_synchronization.cpp
index 04d65f0bd..53df5bcd8 100644
--- a/src/core/hle/kernel/svc/svc_synchronization.cpp
+++ b/src/core/hle/kernel/svc/svc_synchronization.cpp
@@ -2,6 +2,7 @@
 // SPDX-License-Identifier: GPL-2.0-or-later
 #include "common/scope_exit.h"
+#include "common/scratch_buffer.h"
 #include "core/core.h"
 #include "core/hle/kernel/k_process.h"
 #include "core/hle/kernel/k_readable_event.h"
@@ -54,7 +55,7 @@ static Result WaitSynchronization(Core::System& system, int32_t* out_index, cons
    // Get the synchronization context.
    auto& kernel = system.Kernel();
    auto& handle_table = GetCurrentProcess(kernel).GetHandleTable();
-    std::vector<KSynchronizationObject*> objs(num_handles);
+    std::array<KSynchronizationObject*, Svc::ArgumentHandleCountMax> objs;
    // Copy user handles.
    if (num_handles > 0) {
@@ -72,8 +73,8 @@ static Result WaitSynchronization(Core::System& system, int32_t* out_index, cons
    });
    // Wait on the objects.
-    Result res = KSynchronizationObject::Wait(kernel, out_index, objs.data(),
+    Result res =
-                                              static_cast<s32>(objs.size()), timeout_ns);
+        KSynchronizationObject::Wait(kernel, out_index, objs.data(), num_handles, timeout_ns);
    R_SUCCEED_IF(res == ResultSessionClosed);
    R_RETURN(res);
@@ -87,8 +88,7 @@ Result WaitSynchronization(Core::System& system, int32_t* out_index, u64 user_ha
    // Ensure number of handles is valid.
    R_UNLESS(0 <= num_handles && num_handles <= Svc::ArgumentHandleCountMax, ResultOutOfRange);
+    std::array<Handle, Svc::ArgumentHandleCountMax> handles;
-    std::vector<Handle> handles(num_handles);
    if (num_handles > 0) {
        GetCurrentMemory(system.Kernel())
            .ReadBlock(user_handles, handles.data(), num_handles * sizeof(Handle));
diff --git a/src/core/hle/kernel/svc/svc_thread.cpp b/src/core/hle/kernel/svc/svc_thread.cpp
index 37b54079c..36b94e6bf 100644
--- a/src/core/hle/kernel/svc/svc_thread.cpp
+++ b/src/core/hle/kernel/svc/svc_thread.cpp
@@ -174,7 +174,7 @@ Result GetThreadContext3(Core::System& system, u64 out_context, Handle thread_ha
        }
        // Get the thread context.
-        std::vector<u8> context;
+        static thread_local Common::ScratchBuffer<u8> context;
        R_TRY(thread->GetThreadContext3(context));
        // Copy the thread context to user space.
diff --git a/src/core/hle/kernel/svc/svc_tick.cpp b/src/core/hle/kernel/svc/svc_tick.cpp
index 561336482..7dd7c6e51 100644
--- a/src/core/hle/kernel/svc/svc_tick.cpp
+++ b/src/core/hle/kernel/svc/svc_tick.cpp
@@ -12,16 +12,8 @@ namespace Kernel::Svc {
 int64_t GetSystemTick(Core::System& system) {
    LOG_TRACE(Kernel_SVC, "called");
-    auto& core_timing = system.CoreTiming();
    // Returns the value of cntpct_el0 (https://switchbrew.org/wiki/SVC#svcGetSystemTick)
-    const u64 result{core_timing.GetClockTicks()};
+    return static_cast<int64_t>(system.CoreTiming().GetClockTicks());
-    if (!system.Kernel().IsMulticore()) {
-        core_timing.AddTicks(400U);
-    }
-    return static_cast<int64_t>(result);
 }
 int64_t GetSystemTick64(Core::System& system) {
diff --git a/src/core/hle/service/am/applets/applet_cabinet.cpp b/src/core/hle/service/am/applets/applet_cabinet.cpp
index 8b754e9d4..19ed184e8 100644
--- a/src/core/hle/service/am/applets/applet_cabinet.cpp
+++ b/src/core/hle/service/am/applets/applet_cabinet.cpp
@@ -141,7 +141,7 @@ void Cabinet::DisplayCompleted(bool apply_changes, std::string_view amiibo_name)
    applet_output.device_handle = applet_input_common.device_handle;
    applet_output.result = CabinetResult::Cancel;
    const auto reg_result = nfp_device->GetRegisterInfo(applet_output.register_info);
-    const auto tag_result = nfp_device->GetTagInfo(applet_output.tag_info, false);
+    const auto tag_result = nfp_device->GetTagInfo(applet_output.tag_info);
    nfp_device->Finalize();
    if (reg_result.IsSuccess()) {
diff --git a/src/core/hle/service/audio/audin_u.cpp b/src/core/hle/service/audio/audin_u.cpp
index f0640c64f..c8d574993 100644
--- a/src/core/hle/service/audio/audin_u.cpp
+++ b/src/core/hle/service/audio/audin_u.cpp
@@ -5,6 +5,7 @@
 #include "audio_core/renderer/audio_device.h"
 #include "common/common_funcs.h"
 #include "common/logging/log.h"
+#include "common/settings.h"
 #include "common/string_util.h"
 #include "core/core.h"
 #include "core/hle/kernel/k_event.h"
@@ -123,19 +124,13 @@ private:
    void GetReleasedAudioInBuffer(HLERequestContext& ctx) {
        const auto write_buffer_size = ctx.GetWriteBufferNumElements<u64>();
-        std::vector<u64> released_buffers(write_buffer_size);
+        tmp_buffer.resize_destructive(write_buffer_size);
+        tmp_buffer[0] = 0;
-        const auto count = impl->GetReleasedBuffers(released_buffers);
+        const auto count = impl->GetReleasedBuffers(tmp_buffer);
-        [[maybe_unused]] std::string tags{};
+        ctx.WriteBuffer(tmp_buffer);
-        for (u32 i = 0; i < count; i++) {
-            tags += fmt::format("{:08X}, ", released_buffers[i]);
-        }
-        [[maybe_unused]] auto sessionid{impl->GetSystem().GetSessionId()};
-        LOG_TRACE(Service_Audio, "called. Session {} released {} buffers: {}", sessionid, count,
-                  tags);
-        ctx.WriteBuffer(released_buffers);
        IPC::ResponseBuilder rb{ctx, 3};
        rb.Push(ResultSuccess);
        rb.Push(count);
@@ -200,6 +195,7 @@ private:
    KernelHelpers::ServiceContext service_context;
    Kernel::KEvent* event;
    std::shared_ptr<AudioCore::AudioIn::In> impl;
+    Common::ScratchBuffer<u64> tmp_buffer;
 };
 AudInU::AudInU(Core::System& system_)
diff --git a/src/core/hle/service/audio/audout_u.cpp b/src/core/hle/service/audio/audout_u.cpp
index 3e62fa4fc..032c8c11f 100644
--- a/src/core/hle/service/audio/audout_u.cpp
+++ b/src/core/hle/service/audio/audout_u.cpp
@@ -123,19 +123,13 @@ private:
    void GetReleasedAudioOutBuffers(HLERequestContext& ctx) {
        const auto write_buffer_size = ctx.GetWriteBufferNumElements<u64>();
-        std::vector<u64> released_buffers(write_buffer_size);
+        tmp_buffer.resize_destructive(write_buffer_size);
+        tmp_buffer[0] = 0;
-        const auto count = impl->GetReleasedBuffers(released_buffers);
+        const auto count = impl->GetReleasedBuffers(tmp_buffer);
-        [[maybe_unused]] std::string tags{};
+        ctx.WriteBuffer(tmp_buffer);
-        for (u32 i = 0; i < count; i++) {
-            tags += fmt::format("{:08X}, ", released_buffers[i]);
-        }
-        [[maybe_unused]] const auto sessionid{impl->GetSystem().GetSessionId()};
-        LOG_TRACE(Service_Audio, "called. Session {} released {} buffers: {}", sessionid, count,
-                  tags);
-        ctx.WriteBuffer(released_buffers);
        IPC::ResponseBuilder rb{ctx, 3};
        rb.Push(ResultSuccess);
        rb.Push(count);
@@ -211,6 +205,7 @@ private:
    KernelHelpers::ServiceContext service_context;
    Kernel::KEvent* event;
    std::shared_ptr<AudioCore::AudioOut::Out> impl;
+    Common::ScratchBuffer<u64> tmp_buffer;
 };
 AudOutU::AudOutU(Core::System& system_)
diff --git a/src/core/hle/service/audio/audren_u.cpp b/src/core/hle/service/audio/audren_u.cpp
index 7086d4750..12845c23a 100644
--- a/src/core/hle/service/audio/audren_u.cpp
+++ b/src/core/hle/service/audio/audren_u.cpp
@@ -116,28 +116,26 @@ private:
        // These buffers are written manually to avoid an issue with WriteBuffer throwing errors for
        // checking size 0. Performance size is 0 for most games.
-        std::vector<u8> output{};
-        std::vector<u8> performance{};
        auto is_buffer_b{ctx.BufferDescriptorB()[0].Size() != 0};
        if (is_buffer_b) {
            const auto buffersB{ctx.BufferDescriptorB()};
-            output.resize(buffersB[0].Size(), 0);
+            tmp_output.resize_destructive(buffersB[0].Size());
-            performance.resize(buffersB[1].Size(), 0);
+            tmp_performance.resize_destructive(buffersB[1].Size());
        } else {
            const auto buffersC{ctx.BufferDescriptorC()};
-            output.resize(buffersC[0].Size(), 0);
+            tmp_output.resize_destructive(buffersC[0].Size());
-            performance.resize(buffersC[1].Size(), 0);
+            tmp_performance.resize_destructive(buffersC[1].Size());
        }
-        auto result = impl->RequestUpdate(input, performance, output);
+        auto result = impl->RequestUpdate(input, tmp_performance, tmp_output);
        if (result.IsSuccess()) {
            if (is_buffer_b) {
-                ctx.WriteBufferB(output.data(), output.size(), 0);
+                ctx.WriteBufferB(tmp_output.data(), tmp_output.size(), 0);
-                ctx.WriteBufferB(performance.data(), performance.size(), 1);
+                ctx.WriteBufferB(tmp_performance.data(), tmp_performance.size(), 1);
            } else {
-                ctx.WriteBufferC(output.data(), output.size(), 0);
+                ctx.WriteBufferC(tmp_output.data(), tmp_output.size(), 0);
-                ctx.WriteBufferC(performance.data(), performance.size(), 1);
+                ctx.WriteBufferC(tmp_performance.data(), tmp_performance.size(), 1);
            }
        } else {
            LOG_ERROR(Service_Audio, "RequestUpdate failed error 0x{:02X}!", result.description);
@@ -235,6 +233,8 @@ private:
    Kernel::KEvent* rendered_event;
    Manager& manager;
    std::unique_ptr<Renderer> impl;
+    Common::ScratchBuffer<u8> tmp_output;
+    Common::ScratchBuffer<u8> tmp_performance;
 };
 class IAudioDevice final : public ServiceFramework<IAudioDevice> {
diff --git a/src/core/hle/service/audio/audren_u.h b/src/core/hle/service/audio/audren_u.h
index 24ce37e87..d8e9c8719 100644
--- a/src/core/hle/service/audio/audren_u.h
+++ b/src/core/hle/service/audio/audren_u.h
@@ -4,6 +4,7 @@
 #pragma once
 #include "audio_core/audio_render_manager.h"
+#include "common/scratch_buffer.h"
 #include "core/hle/service/kernel_helpers.h"
 #include "core/hle/service/service.h"
diff --git a/src/core/hle/service/audio/hwopus.cpp b/src/core/hle/service/audio/hwopus.cpp
index 451ac224a..c835f6cb7 100644
--- a/src/core/hle/service/audio/hwopus.cpp
+++ b/src/core/hle/service/audio/hwopus.cpp
@@ -68,13 +68,13 @@ private:
                                 ExtraBehavior extra_behavior) {
        u32 consumed = 0;
        u32 sample_count = 0;
-        std::vector<opus_int16> samples(ctx.GetWriteBufferNumElements<opus_int16>());
+        tmp_samples.resize_destructive(ctx.GetWriteBufferNumElements<opus_int16>());
        if (extra_behavior == ExtraBehavior::ResetContext) {
            ResetDecoderContext();
        }
-        if (!DecodeOpusData(consumed, sample_count, ctx.ReadBuffer(), samples, performance)) {
+        if (!DecodeOpusData(consumed, sample_count, ctx.ReadBuffer(), tmp_samples, performance)) {
            LOG_ERROR(Audio, "Failed to decode opus data");
            IPC::ResponseBuilder rb{ctx, 2};
            // TODO(ogniK): Use correct error code
@@ -90,11 +90,11 @@ private:
        if (performance) {
            rb.Push<u64>(*performance);
        }
-        ctx.WriteBuffer(samples);
+        ctx.WriteBuffer(tmp_samples);
    }
    bool DecodeOpusData(u32& consumed, u32& sample_count, std::span<const u8> input,
-                        std::vector<opus_int16>& output, u64* out_performance_time) const {
+                        std::span<opus_int16> output, u64* out_performance_time) const {
        const auto start_time = std::chrono::steady_clock::now();
        const std::size_t raw_output_sz = output.size() * sizeof(opus_int16);
        if (sizeof(OpusPacketHeader) > input.size()) {
@@ -154,6 +154,7 @@ private:
    OpusDecoderPtr decoder;
    u32 sample_rate;
    u32 channel_count;
+    Common::ScratchBuffer<opus_int16> tmp_samples;
 };
 class IHardwareOpusDecoderManager final : public ServiceFramework<IHardwareOpusDecoderManager> {
diff --git a/src/core/hle/service/hid/hidbus.cpp b/src/core/hle/service/hid/hidbus.cpp
index 5604a6fda..80aac221b 100644
--- a/src/core/hle/service/hid/hidbus.cpp
+++ b/src/core/hle/service/hid/hidbus.cpp
@@ -5,7 +5,6 @@
 #include "common/settings.h"
 #include "core/core.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 #include "core/hid/hid_types.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/kernel/k_readable_event.h"
diff --git a/src/core/hle/service/nfc/common/device.cpp b/src/core/hle/service/nfc/common/device.cpp
index f4b180b06..5bf289818 100644
--- a/src/core/hle/service/nfc/common/device.cpp
+++ b/src/core/hle/service/nfc/common/device.cpp
@@ -93,7 +93,8 @@ void NfcDevice::NpadUpdate(Core::HID::ControllerTriggerType type) {
    const auto nfc_status = npad_device->GetNfc();
    switch (nfc_status.state) {
    case Common::Input::NfcState::NewAmiibo:
-        LoadNfcTag(nfc_status.data);
+        LoadNfcTag(nfc_status.protocol, nfc_status.tag_type, nfc_status.uuid_length,
+                   nfc_status.uuid);
        break;
    case Common::Input::NfcState::AmiiboRemoved:
        if (device_state == DeviceState::Initialized || device_state == DeviceState::TagRemoved) {
@@ -108,28 +109,46 @@ void NfcDevice::NpadUpdate(Core::HID::ControllerTriggerType type) {
    }
 }
-bool NfcDevice::LoadNfcTag(std::span<const u8> data) {
+bool NfcDevice::LoadNfcTag(u8 protocol, u8 tag_type, u8 uuid_length, UniqueSerialNumber uuid) {
    if (device_state != DeviceState::SearchingForTag) {
        LOG_ERROR(Service_NFC, "Game is not looking for nfc tag, current state {}", device_state);
        return false;
    }
+    if ((protocol & static_cast<u8>(allowed_protocols)) == 0) {
+        LOG_ERROR(Service_NFC, "Protocol not supported {}", protocol);
+        return false;
+    }
+    real_tag_info = {
+        .uuid = uuid,
+        .uuid_length = uuid_length,
+        .protocol = static_cast<NfcProtocol>(protocol),
+        .tag_type = static_cast<TagType>(tag_type),
+    };
+    device_state = DeviceState::TagFound;
+    deactivate_event->GetReadableEvent().Clear();
+    activate_event->Signal();
+    return true;
+}
+bool NfcDevice::LoadAmiiboData() {
+    std::vector<u8> data{};
+    if (!npad_device->ReadAmiiboData(data)) {
+        return false;
+    }
    if (data.size() < sizeof(NFP::EncryptedNTAG215File)) {
        LOG_ERROR(Service_NFC, "Not an amiibo, size={}", data.size());
        return false;
    }
-    mifare_data.resize(data.size());
-    memcpy(mifare_data.data(), data.data(), data.size());
    memcpy(&tag_data, data.data(), sizeof(NFP::EncryptedNTAG215File));
    is_plain_amiibo = NFP::AmiiboCrypto::IsAmiiboValid(tag_data);
    is_write_protected = false;
-    device_state = DeviceState::TagFound;
-    deactivate_event->GetReadableEvent().Clear();
-    activate_event->Signal();
    // Fallback for plain amiibos
    if (is_plain_amiibo) {
        LOG_INFO(Service_NFP, "Using plain amiibo");
@@ -147,6 +166,7 @@ bool NfcDevice::LoadNfcTag(std::span<const u8> data) {
        return true;
    }
+    LOG_INFO(Service_NFP, "Using encrypted amiibo");
    tag_data = {};
    memcpy(&encrypted_tag_data, data.data(), sizeof(NFP::EncryptedNTAG215File));
    return true;
@@ -162,7 +182,6 @@ void NfcDevice::CloseNfcTag() {
    device_state = DeviceState::TagRemoved;
    encrypted_tag_data = {};
    tag_data = {};
-    mifare_data = {};
    activate_event->GetReadableEvent().Clear();
    deactivate_event->Signal();
 }
@@ -179,8 +198,12 @@ void NfcDevice::Initialize() {
    device_state = npad_device->HasNfc() ? DeviceState::Initialized : DeviceState::Unavailable;
    encrypted_tag_data = {};
    tag_data = {};
-    mifare_data = {};
-    is_initalized = true;
+    if (device_state != DeviceState::Initialized) {
+        return;
+    }
+    is_initalized = npad_device->AddNfcHandle();
 }
 void NfcDevice::Finalize() {
@@ -190,6 +213,11 @@ void NfcDevice::Finalize() {
    if (device_state == DeviceState::SearchingForTag || device_state == DeviceState::TagRemoved) {
        StopDetection();
    }
+    if (device_state != DeviceState::Unavailable) {
+        npad_device->RemoveNfcHandle();
+    }
    device_state = DeviceState::Unavailable;
    is_initalized = false;
 }
@@ -200,10 +228,8 @@ Result NfcDevice::StartDetection(NfcProtocol allowed_protocol) {
        return ResultWrongDeviceState;
    }
-    if (npad_device->SetPollingMode(Core::HID::EmulatedDeviceIndex::RightIndex,
+    if (!npad_device->StartNfcPolling()) {
-                                    Common::Input::PollingMode::NFC) !=
+        LOG_ERROR(Service_NFC, "Nfc polling not supported");
-        Common::Input::DriverResult::Success) {
-        LOG_ERROR(Service_NFC, "Nfc not supported");
        return ResultNfcDisabled;
    }
@@ -213,9 +239,6 @@ Result NfcDevice::StartDetection(NfcProtocol allowed_protocol) {
 }
 Result NfcDevice::StopDetection() {
-    npad_device->SetPollingMode(Core::HID::EmulatedDeviceIndex::RightIndex,
-                                Common::Input::PollingMode::Active);
    if (device_state == DeviceState::Initialized) {
        return ResultSuccess;
    }
@@ -225,6 +248,7 @@ Result NfcDevice::StopDetection() {
    }
    if (device_state == DeviceState::SearchingForTag || device_state == DeviceState::TagRemoved) {
+        npad_device->StopNfcPolling();
        device_state = DeviceState::Initialized;
        return ResultSuccess;
    }
@@ -233,7 +257,7 @@ Result NfcDevice::StopDetection() {
    return ResultWrongDeviceState;
 }
-Result NfcDevice::GetTagInfo(NFP::TagInfo& tag_info, bool is_mifare) const {
+Result NfcDevice::GetTagInfo(NFP::TagInfo& tag_info) const {
    if (device_state != DeviceState::TagFound && device_state != DeviceState::TagMounted) {
        LOG_ERROR(Service_NFC, "Wrong device state {}", device_state);
        if (device_state == DeviceState::TagRemoved) {
@@ -242,41 +266,15 @@ Result NfcDevice::GetTagInfo(NFP::TagInfo& tag_info, bool is_mifare) const {
        return ResultWrongDeviceState;
    }
-    UniqueSerialNumber uuid{};
+    tag_info = real_tag_info;
-    u8 uuid_length{};
-    NfcProtocol protocol{NfcProtocol::TypeA};
-    TagType tag_type{TagType::Type2};
-    if (is_mifare) {
-        tag_type = TagType::Mifare;
-        uuid_length = sizeof(NFP::NtagTagUuid);
-        memcpy(uuid.data(), mifare_data.data(), uuid_length);
-    } else {
-        tag_type = TagType::Type2;
-        uuid_length = sizeof(NFP::NtagTagUuid);
-        NFP::NtagTagUuid nUuid{
-            .part1 = encrypted_tag_data.uuid.part1,
-            .part2 = encrypted_tag_data.uuid.part2,
-            .nintendo_id = encrypted_tag_data.uuid.nintendo_id,
-        };
-        memcpy(uuid.data(), &nUuid, uuid_length);
-        // Generate random UUID to bypass amiibo load limits
+    // Generate random UUID to bypass amiibo load limits
-        if (Settings::values.random_amiibo_id) {
+    if (real_tag_info.tag_type == TagType::Type2 && Settings::values.random_amiibo_id) {
-            Common::TinyMT rng{};
+        Common::TinyMT rng{};
-            rng.Initialize(static_cast<u32>(GetCurrentPosixTime()));
+        rng.Initialize(static_cast<u32>(GetCurrentPosixTime()));
-            rng.GenerateRandomBytes(uuid.data(), uuid_length);
+        rng.GenerateRandomBytes(tag_info.uuid.data(), tag_info.uuid_length);
-        }
    }
-    // Protocol and tag type may change here
-    tag_info = {
-        .uuid = uuid,
-        .uuid_length = uuid_length,
-        .protocol = protocol,
-        .tag_type = tag_type,
-    };
    return ResultSuccess;
 }
@@ -293,7 +291,7 @@ Result NfcDevice::ReadMifare(std::span<const MifareReadBlockParameter> parameter
    Result result = ResultSuccess;
    TagInfo tag_info{};
-    result = GetTagInfo(tag_info, true);
+    result = GetTagInfo(tag_info);
    if (result.IsError()) {
        return result;
@@ -307,6 +305,8 @@ Result NfcDevice::ReadMifare(std::span<const MifareReadBlockParameter> parameter
        return ResultInvalidArgument;
    }
+    Common::Input::MifareRequest request{};
+    Common::Input::MifareRequest out_data{};
    const auto unknown = parameters[0].sector_key.unknown;
    for (std::size_t i = 0; i < parameters.size(); i++) {
        if (unknown != parameters[i].sector_key.unknown) {
@@ -315,25 +315,29 @@ Result NfcDevice::ReadMifare(std::span<const MifareReadBlockParameter> parameter
    }
    for (std::size_t i = 0; i < parameters.size(); i++) {
-        result = ReadMifare(parameters[i], read_block_data[i]);
+        if (parameters[i].sector_key.command == MifareCmd::None) {
-        if (result.IsError()) {
+            continue;
-            break;
        }
+        request.data[i].command = static_cast<u8>(parameters[i].sector_key.command);
+        request.data[i].sector = parameters[i].sector_number;
+        memcpy(request.data[i].key.data(), parameters[i].sector_key.sector_key.data(),
+               sizeof(KeyData));
    }
-    return result;
+    if (!npad_device->ReadMifareData(request, out_data)) {
-}
-Result NfcDevice::ReadMifare(const MifareReadBlockParameter& parameter,
-                             MifareReadBlockData& read_block_data) const {
-    const std::size_t sector_index = parameter.sector_number * sizeof(DataBlock);
-    read_block_data.sector_number = parameter.sector_number;
-    if (mifare_data.size() < sector_index + sizeof(DataBlock)) {
        return ResultMifareError288;
    }
-    // TODO: Use parameter.sector_key to read encrypted data
+    for (std::size_t i = 0; i < read_block_data.size(); i++) {
-    memcpy(read_block_data.data.data(), mifare_data.data() + sector_index, sizeof(DataBlock));
+        if (static_cast<MifareCmd>(out_data.data[i].command) == MifareCmd::None) {
+            continue;
+        }
+        read_block_data[i] = {
+            .data = out_data.data[i].data,
+            .sector_number = out_data.data[i].sector,
+        };
+    }
    return ResultSuccess;
 }
@@ -342,7 +346,7 @@ Result NfcDevice::WriteMifare(std::span<const MifareWriteBlockParameter> paramet
    Result result = ResultSuccess;
    TagInfo tag_info{};
-    result = GetTagInfo(tag_info, true);
+    result = GetTagInfo(tag_info);
    if (result.IsError()) {
        return result;
@@ -363,42 +367,25 @@ Result NfcDevice::WriteMifare(std::span<const MifareWriteBlockParameter> paramet
        }
    }
+    Common::Input::MifareRequest request{};
    for (std::size_t i = 0; i < parameters.size(); i++) {
-        result = WriteMifare(parameters[i]);
+        if (parameters[i].sector_key.command == MifareCmd::None) {
-        if (result.IsError()) {
+            continue;
-            break;
        }
+        request.data[i].command = static_cast<u8>(parameters[i].sector_key.command);
+        request.data[i].sector = parameters[i].sector_number;
+        memcpy(request.data[i].key.data(), parameters[i].sector_key.sector_key.data(),
+               sizeof(KeyData));
+        memcpy(request.data[i].data.data(), parameters[i].data.data(), sizeof(KeyData));
    }
-    if (!npad_device->WriteNfc(mifare_data)) {
+    if (!npad_device->WriteMifareData(request)) {
-        LOG_ERROR(Service_NFP, "Error writing to file");
        return ResultMifareError288;
    }
    return result;
 }
-Result NfcDevice::WriteMifare(const MifareWriteBlockParameter& parameter) {
-    const std::size_t sector_index = parameter.sector_number * sizeof(DataBlock);
-    if (device_state != DeviceState::TagFound && device_state != DeviceState::TagMounted) {
-        LOG_ERROR(Service_NFC, "Wrong device state {}", device_state);
-        if (device_state == DeviceState::TagRemoved) {
-            return ResultTagRemoved;
-        }
-        return ResultWrongDeviceState;
-    }
-    if (mifare_data.size() < sector_index + sizeof(DataBlock)) {
-        return ResultMifareError288;
-    }
-    // TODO: Use parameter.sector_key to encrypt the data
-    memcpy(mifare_data.data() + sector_index, parameter.data.data(), sizeof(DataBlock));
-    return ResultSuccess;
-}
 Result NfcDevice::SendCommandByPassThrough(const Time::Clock::TimeSpanType& timeout,
                                           std::span<const u8> command_data,
                                           std::span<u8> out_data) {
@@ -412,6 +399,11 @@ Result NfcDevice::Mount(NFP::ModelType model_type, NFP::MountTarget mount_target
        return ResultWrongDeviceState;
    }
+    if (!LoadAmiiboData()) {
+        LOG_ERROR(Service_NFP, "Not an amiibo");
+        return ResultInvalidTagType;
+    }
    if (!NFP::AmiiboCrypto::IsAmiiboValid(encrypted_tag_data)) {
        LOG_ERROR(Service_NFP, "Not an amiibo");
        return ResultInvalidTagType;
@@ -562,7 +554,7 @@ Result NfcDevice::Restore() {
    NFC::TagInfo tag_info{};
    std::array<u8, sizeof(NFP::EncryptedNTAG215File)> data{};
-    Result result = GetTagInfo(tag_info, false);
+    Result result = GetTagInfo(tag_info);
    if (result.IsError()) {
        return result;
@@ -635,7 +627,7 @@ Result NfcDevice::GetCommonInfo(NFP::CommonInfo& common_info) const {
    // TODO: Validate this data
    common_info = {
        .last_write_date = settings.write_date.GetWriteDate(),
-        .write_counter = tag_data.write_counter,
+        .write_counter = tag_data.application_write_counter,
        .version = tag_data.amiibo_version,
        .application_area_size = sizeof(NFP::ApplicationArea),
    };
diff --git a/src/core/hle/service/nfc/common/device.h b/src/core/hle/service/nfc/common/device.h
index 7560210d6..0ed1ff34c 100644
--- a/src/core/hle/service/nfc/common/device.h
+++ b/src/core/hle/service/nfc/common/device.h
@@ -42,15 +42,12 @@ public:
    Result StartDetection(NfcProtocol allowed_protocol);
    Result StopDetection();
-    Result GetTagInfo(TagInfo& tag_info, bool is_mifare) const;
+    Result GetTagInfo(TagInfo& tag_info) const;
    Result ReadMifare(std::span<const MifareReadBlockParameter> parameters,
                      std::span<MifareReadBlockData> read_block_data) const;
-    Result ReadMifare(const MifareReadBlockParameter& parameter,
-                      MifareReadBlockData& read_block_data) const;
    Result WriteMifare(std::span<const MifareWriteBlockParameter> parameters);
-    Result WriteMifare(const MifareWriteBlockParameter& parameter);
    Result SendCommandByPassThrough(const Time::Clock::TimeSpanType& timeout,
                                    std::span<const u8> command_data, std::span<u8> out_data);
@@ -105,7 +102,8 @@ public:
 private:
    void NpadUpdate(Core::HID::ControllerTriggerType type);
-    bool LoadNfcTag(std::span<const u8> data);
+    bool LoadNfcTag(u8 protocol, u8 tag_type, u8 uuid_length, UniqueSerialNumber uuid);
+    bool LoadAmiiboData();
    void CloseNfcTag();
    NFP::AmiiboName GetAmiiboName(const NFP::AmiiboSettings& settings) const;
@@ -140,8 +138,8 @@ private:
    bool is_write_protected{};
    NFP::MountTarget mount_target{NFP::MountTarget::None};
+    TagInfo real_tag_info{};
    NFP::NTAG215File tag_data{};
-    std::vector<u8> mifare_data{};
    NFP::EncryptedNTAG215File encrypted_tag_data{};
 };
diff --git a/src/core/hle/service/nfc/common/device_manager.cpp b/src/core/hle/service/nfc/common/device_manager.cpp
index b0456508e..562f3a28e 100644
--- a/src/core/hle/service/nfc/common/device_manager.cpp
+++ b/src/core/hle/service/nfc/common/device_manager.cpp
@@ -29,6 +29,9 @@ DeviceManager::DeviceManager(Core::System& system_, KernelHelpers::ServiceContex
 }
 DeviceManager ::~DeviceManager() {
+    if (is_initialized) {
+        Finalize();
+    }
    service_context.CloseEvent(availability_change_event);
 }
@@ -125,14 +128,14 @@ Result DeviceManager::StopDetection(u64 device_handle) {
    return result;
 }
-Result DeviceManager::GetTagInfo(u64 device_handle, TagInfo& tag_info, bool is_mifare) const {
+Result DeviceManager::GetTagInfo(u64 device_handle, TagInfo& tag_info) const {
    std::scoped_lock lock{mutex};
    std::shared_ptr<NfcDevice> device = nullptr;
    auto result = GetDeviceHandle(device_handle, device);
    if (result.IsSuccess()) {
-        result = device->GetTagInfo(tag_info, is_mifare);
+        result = device->GetTagInfo(tag_info);
        result = VerifyDeviceResult(device, result);
    }
@@ -546,7 +549,7 @@ Result DeviceManager::ReadBackupData(u64 device_handle, std::span<u8> data) cons
    NFC::TagInfo tag_info{};
    if (result.IsSuccess()) {
-        result = device->GetTagInfo(tag_info, false);
+        result = device->GetTagInfo(tag_info);
    }
    if (result.IsSuccess()) {
@@ -565,7 +568,7 @@ Result DeviceManager::WriteBackupData(u64 device_handle, std::span<const u8> dat
    NFC::TagInfo tag_info{};
    if (result.IsSuccess()) {
-        result = device->GetTagInfo(tag_info, false);
+        result = device->GetTagInfo(tag_info);
    }
    if (result.IsSuccess()) {
diff --git a/src/core/hle/service/nfc/common/device_manager.h b/src/core/hle/service/nfc/common/device_manager.h
index 2971e280f..c61ba0cf3 100644
--- a/src/core/hle/service/nfc/common/device_manager.h
+++ b/src/core/hle/service/nfc/common/device_manager.h
@@ -33,7 +33,7 @@ public:
    Kernel::KReadableEvent& AttachAvailabilityChangeEvent() const;
    Result StartDetection(u64 device_handle, NfcProtocol tag_protocol);
    Result StopDetection(u64 device_handle);
-    Result GetTagInfo(u64 device_handle, NFP::TagInfo& tag_info, bool is_mifare) const;
+    Result GetTagInfo(u64 device_handle, NFP::TagInfo& tag_info) const;
    Kernel::KReadableEvent& AttachActivateEvent(u64 device_handle) const;
    Kernel::KReadableEvent& AttachDeactivateEvent(u64 device_handle) const;
    Result ReadMifare(u64 device_handle,
diff --git a/src/core/hle/service/nfc/mifare_types.h b/src/core/hle/service/nfc/mifare_types.h
index 75b59f021..467937399 100644
--- a/src/core/hle/service/nfc/mifare_types.h
+++ b/src/core/hle/service/nfc/mifare_types.h
@@ -11,9 +11,10 @@
 namespace Service::NFC {
 enum class MifareCmd : u8 {
+    None = 0x00,
+    Read = 0x30,
    AuthA = 0x60,
    AuthB = 0x61,
-    Read = 0x30,
    Write = 0xA0,
    Transfer = 0xB0,
    Decrement = 0xC0,
@@ -35,17 +36,17 @@ static_assert(sizeof(SectorKey) == 0x10, "SectorKey is an invalid size");
 // This is nn::nfc::MifareReadBlockParameter
 struct MifareReadBlockParameter {
-    u8 sector_number;
+    u8 sector_number{};
    INSERT_PADDING_BYTES(0x7);
-    SectorKey sector_key;
+    SectorKey sector_key{};
 };
 static_assert(sizeof(MifareReadBlockParameter) == 0x18,
              "MifareReadBlockParameter is an invalid size");
 // This is nn::nfc::MifareReadBlockData
 struct MifareReadBlockData {
-    DataBlock data;
+    DataBlock data{};
-    u8 sector_number;
+    u8 sector_number{};
    INSERT_PADDING_BYTES(0x7);
 };
 static_assert(sizeof(MifareReadBlockData) == 0x18, "MifareReadBlockData is an invalid size");
diff --git a/src/core/hle/service/nfc/nfc_interface.cpp b/src/core/hle/service/nfc/nfc_interface.cpp
index 130fb7f78..e7ca7582e 100644
--- a/src/core/hle/service/nfc/nfc_interface.cpp
+++ b/src/core/hle/service/nfc/nfc_interface.cpp
@@ -174,8 +174,7 @@ void NfcInterface::GetTagInfo(HLERequestContext& ctx) {
    LOG_INFO(Service_NFC, "called, device_handle={}", device_handle);
    TagInfo tag_info{};
-    auto result =
+    auto result = GetManager()->GetTagInfo(device_handle, tag_info);
-        GetManager()->GetTagInfo(device_handle, tag_info, backend_type == BackendType::Mifare);
    result = TranslateResultToServiceError(result);
    if (result.IsSuccess()) {
@@ -216,8 +215,8 @@ void NfcInterface::ReadMifare(HLERequestContext& ctx) {
    memcpy(read_commands.data(), buffer.data(),
           number_of_commands * sizeof(MifareReadBlockParameter));
-    LOG_INFO(Service_NFC, "(STUBBED) called, device_handle={}, read_commands_size={}",
+    LOG_INFO(Service_NFC, "called, device_handle={}, read_commands_size={}", device_handle,
-             device_handle, number_of_commands);
+             number_of_commands);
    std::vector<MifareReadBlockData> out_data(number_of_commands);
    auto result = GetManager()->ReadMifare(device_handle, read_commands, out_data);
diff --git a/src/core/hle/service/nvdrv/devices/nvdevice.h b/src/core/hle/service/nvdrv/devices/nvdevice.h
index ab1f30f9e..a04538d5d 100644
--- a/src/core/hle/service/nvdrv/devices/nvdevice.h
+++ b/src/core/hle/service/nvdrv/devices/nvdevice.h
@@ -34,7 +34,7 @@ public:
     * @returns The result code of the ioctl.
     */
    virtual NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::vector<u8>& output) = 0;
+                            std::span<u8> output) = 0;
    /**
     * Handles an ioctl2 request.
@@ -45,7 +45,7 @@ public:
     * @returns The result code of the ioctl.
     */
    virtual NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::span<const u8> inline_input, std::vector<u8>& output) = 0;
+                            std::span<const u8> inline_input, std::span<u8> output) = 0;
    /**
     * Handles an ioctl3 request.
@@ -56,7 +56,7 @@ public:
     * @returns The result code of the ioctl.
     */
    virtual NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::vector<u8>& output, std::vector<u8>& inline_output) = 0;
+                            std::span<u8> output, std::span<u8> inline_output) = 0;
    /**
     * Called once a device is opened
diff --git a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp
index 5a5b2e305..05a43d8dc 100644
--- a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp
@@ -18,19 +18,19 @@ nvdisp_disp0::nvdisp_disp0(Core::System& system_, NvCore::Container& core)
 nvdisp_disp0::~nvdisp_disp0() = default;
 NvResult nvdisp_disp0::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::vector<u8>& output) {
+                              std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
 NvResult nvdisp_disp0::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::span<const u8> inline_input, std::vector<u8>& output) {
+                              std::span<const u8> inline_input, std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
 NvResult nvdisp_disp0::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::vector<u8>& output, std::vector<u8>& inline_output) {
+                              std::span<u8> output, std::span<u8> inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
@@ -51,8 +51,8 @@ void nvdisp_disp0::flip(u32 buffer_handle, u32 offset, android::PixelFormat form
                                               stride, format, transform, crop_rect};
    system.GPU().RequestSwapBuffers(&framebuffer, fences, num_fences);
-    system.GetPerfStats().EndSystemFrame();
    system.SpeedLimiter().DoSpeedLimiting(system.CoreTiming().GetGlobalTimeUs());
+    system.GetPerfStats().EndSystemFrame();
    system.GetPerfStats().BeginSystemFrame();
 }
diff --git a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h
index bcd0e3ed5..daee05fe8 100644
--- a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h
+++ b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h
@@ -26,11 +26,11 @@ public:
    ~nvdisp_disp0() override;
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.cpp b/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.cpp
index 681bd0867..07e570a9f 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.cpp
@@ -28,7 +28,7 @@ nvhost_as_gpu::nvhost_as_gpu(Core::System& system_, Module& module_, NvCore::Con
 nvhost_as_gpu::~nvhost_as_gpu() = default;
 NvResult nvhost_as_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                               std::vector<u8>& output) {
+                               std::span<u8> output) {
    switch (command.group) {
    case 'A':
        switch (command.cmd) {
@@ -61,13 +61,13 @@ NvResult nvhost_as_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> i
 }
 NvResult nvhost_as_gpu::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                               std::span<const u8> inline_input, std::vector<u8>& output) {
+                               std::span<const u8> inline_input, std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
 NvResult nvhost_as_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                               std::vector<u8>& output, std::vector<u8>& inline_output) {
+                               std::span<u8> output, std::span<u8> inline_output) {
    switch (command.group) {
    case 'A':
        switch (command.cmd) {
@@ -87,7 +87,7 @@ NvResult nvhost_as_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> i
 void nvhost_as_gpu::OnOpen(DeviceFD fd) {}
 void nvhost_as_gpu::OnClose(DeviceFD fd) {}
-NvResult nvhost_as_gpu::AllocAsEx(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::AllocAsEx(std::span<const u8> input, std::span<u8> output) {
    IoctlAllocAsEx params{};
    std::memcpy(&params, input.data(), input.size());
@@ -141,7 +141,7 @@ NvResult nvhost_as_gpu::AllocAsEx(std::span<const u8> input, std::vector<u8>& ou
    return NvResult::Success;
 }
-NvResult nvhost_as_gpu::AllocateSpace(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::AllocateSpace(std::span<const u8> input, std::span<u8> output) {
    IoctlAllocSpace params{};
    std::memcpy(&params, input.data(), input.size());
@@ -220,7 +220,7 @@ void nvhost_as_gpu::FreeMappingLocked(u64 offset) {
    mapping_map.erase(offset);
 }
-NvResult nvhost_as_gpu::FreeSpace(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::FreeSpace(std::span<const u8> input, std::span<u8> output) {
    IoctlFreeSpace params{};
    std::memcpy(&params, input.data(), input.size());
@@ -266,15 +266,14 @@ NvResult nvhost_as_gpu::FreeSpace(std::span<const u8> input, std::vector<u8>& ou
    return NvResult::Success;
 }
-NvResult nvhost_as_gpu::Remap(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::Remap(std::span<const u8> input, std::span<u8> output) {
    const auto num_entries = input.size() / sizeof(IoctlRemapEntry);
    LOG_DEBUG(Service_NVDRV, "called, num_entries=0x{:X}", num_entries);
-    std::vector<IoctlRemapEntry> entries(num_entries);
-    std::memcpy(entries.data(), input.data(), input.size());
    std::scoped_lock lock(mutex);
+    entries.resize_destructive(num_entries);
+    std::memcpy(entries.data(), input.data(), input.size());
    if (!vm.initialised) {
        return NvResult::BadValue;
@@ -320,7 +319,7 @@ NvResult nvhost_as_gpu::Remap(std::span<const u8> input, std::vector<u8>& output
    return NvResult::Success;
 }
-NvResult nvhost_as_gpu::MapBufferEx(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::MapBufferEx(std::span<const u8> input, std::span<u8> output) {
    IoctlMapBufferEx params{};
    std::memcpy(&params, input.data(), input.size());
@@ -424,7 +423,7 @@ NvResult nvhost_as_gpu::MapBufferEx(std::span<const u8> input, std::vector<u8>&
    return NvResult::Success;
 }
-NvResult nvhost_as_gpu::UnmapBuffer(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::UnmapBuffer(std::span<const u8> input, std::span<u8> output) {
    IoctlUnmapBuffer params{};
    std::memcpy(&params, input.data(), input.size());
@@ -463,7 +462,7 @@ NvResult nvhost_as_gpu::UnmapBuffer(std::span<const u8> input, std::vector<u8>&
    return NvResult::Success;
 }
-NvResult nvhost_as_gpu::BindChannel(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::BindChannel(std::span<const u8> input, std::span<u8> output) {
    IoctlBindChannel params{};
    std::memcpy(&params, input.data(), input.size());
    LOG_DEBUG(Service_NVDRV, "called, fd={:X}", params.fd);
@@ -492,7 +491,7 @@ void nvhost_as_gpu::GetVARegionsImpl(IoctlGetVaRegions& params) {
    };
 }
-NvResult nvhost_as_gpu::GetVARegions(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::GetVARegions(std::span<const u8> input, std::span<u8> output) {
    IoctlGetVaRegions params{};
    std::memcpy(&params, input.data(), input.size());
@@ -511,8 +510,8 @@ NvResult nvhost_as_gpu::GetVARegions(std::span<const u8> input, std::vector<u8>&
    return NvResult::Success;
 }
-NvResult nvhost_as_gpu::GetVARegions(std::span<const u8> input, std::vector<u8>& output,
+NvResult nvhost_as_gpu::GetVARegions(std::span<const u8> input, std::span<u8> output,
-                                     std::vector<u8>& inline_output) {
+                                     std::span<u8> inline_output) {
    IoctlGetVaRegions params{};
    std::memcpy(&params, input.data(), input.size());
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.h b/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.h
index 1aba8d579..2af3e1260 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.h
@@ -15,6 +15,7 @@
 #include "common/address_space.h"
 #include "common/common_funcs.h"
 #include "common/common_types.h"
+#include "common/scratch_buffer.h"
 #include "common/swap.h"
 #include "core/hle/service/nvdrv/core/nvmap.h"
 #include "core/hle/service/nvdrv/devices/nvdevice.h"
@@ -48,11 +49,11 @@ public:
    ~nvhost_as_gpu() override;
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
@@ -138,18 +139,18 @@ private:
    static_assert(sizeof(IoctlGetVaRegions) == 16 + sizeof(VaRegion) * 2,
                  "IoctlGetVaRegions is incorrect size");
-    NvResult AllocAsEx(std::span<const u8> input, std::vector<u8>& output);
+    NvResult AllocAsEx(std::span<const u8> input, std::span<u8> output);
-    NvResult AllocateSpace(std::span<const u8> input, std::vector<u8>& output);
+    NvResult AllocateSpace(std::span<const u8> input, std::span<u8> output);
-    NvResult Remap(std::span<const u8> input, std::vector<u8>& output);
+    NvResult Remap(std::span<const u8> input, std::span<u8> output);
-    NvResult MapBufferEx(std::span<const u8> input, std::vector<u8>& output);
+    NvResult MapBufferEx(std::span<const u8> input, std::span<u8> output);
-    NvResult UnmapBuffer(std::span<const u8> input, std::vector<u8>& output);
+    NvResult UnmapBuffer(std::span<const u8> input, std::span<u8> output);
-    NvResult FreeSpace(std::span<const u8> input, std::vector<u8>& output);
+    NvResult FreeSpace(std::span<const u8> input, std::span<u8> output);
-    NvResult BindChannel(std::span<const u8> input, std::vector<u8>& output);
+    NvResult BindChannel(std::span<const u8> input, std::span<u8> output);
    void GetVARegionsImpl(IoctlGetVaRegions& params);
-    NvResult GetVARegions(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetVARegions(std::span<const u8> input, std::span<u8> output);
-    NvResult GetVARegions(std::span<const u8> input, std::vector<u8>& output,
+    NvResult GetVARegions(std::span<const u8> input, std::span<u8> output,
-                          std::vector<u8>& inline_output);
+                          std::span<u8> inline_output);
    void FreeMappingLocked(u64 offset);
@@ -212,6 +213,7 @@ private:
        bool initialised{};
    } vm;
    std::shared_ptr<Tegra::MemoryManager> gmmu;
+    Common::ScratchBuffer<IoctlRemapEntry> entries;
    // s32 channel{};
    // u32 big_page_size{VM::DEFAULT_BIG_PAGE_SIZE};
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl.cpp b/src/core/hle/service/nvdrv/devices/nvhost_ctrl.cpp
index e12025560..4d55554b4 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl.cpp
@@ -35,7 +35,7 @@ nvhost_ctrl::~nvhost_ctrl() {
 }
 NvResult nvhost_ctrl::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                             std::vector<u8>& output) {
+                             std::span<u8> output) {
    switch (command.group) {
    case 0x0:
        switch (command.cmd) {
@@ -64,13 +64,13 @@ NvResult nvhost_ctrl::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> inp
 }
 NvResult nvhost_ctrl::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                             std::span<const u8> inline_input, std::vector<u8>& output) {
+                             std::span<const u8> inline_input, std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
 NvResult nvhost_ctrl::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                             std::vector<u8>& output, std::vector<u8>& inline_outpu) {
+                             std::span<u8> output, std::span<u8> inline_outpu) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
@@ -79,7 +79,7 @@ void nvhost_ctrl::OnOpen(DeviceFD fd) {}
 void nvhost_ctrl::OnClose(DeviceFD fd) {}
-NvResult nvhost_ctrl::NvOsGetConfigU32(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::NvOsGetConfigU32(std::span<const u8> input, std::span<u8> output) {
    IocGetConfigParams params{};
    std::memcpy(&params, input.data(), sizeof(params));
    LOG_TRACE(Service_NVDRV, "called, setting={}!{}", params.domain_str.data(),
@@ -87,7 +87,7 @@ NvResult nvhost_ctrl::NvOsGetConfigU32(std::span<const u8> input, std::vector<u8
    return NvResult::ConfigVarNotFound; // Returns error on production mode
 }
-NvResult nvhost_ctrl::IocCtrlEventWait(std::span<const u8> input, std::vector<u8>& output,
+NvResult nvhost_ctrl::IocCtrlEventWait(std::span<const u8> input, std::span<u8> output,
                                       bool is_allocation) {
    IocCtrlEventWaitParams params{};
    std::memcpy(&params, input.data(), sizeof(params));
@@ -231,7 +231,7 @@ NvResult nvhost_ctrl::FreeEvent(u32 slot) {
    return NvResult::Success;
 }
-NvResult nvhost_ctrl::IocCtrlEventRegister(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::IocCtrlEventRegister(std::span<const u8> input, std::span<u8> output) {
    IocCtrlEventRegisterParams params{};
    std::memcpy(&params, input.data(), sizeof(params));
    const u32 event_id = params.user_event_id;
@@ -252,7 +252,7 @@ NvResult nvhost_ctrl::IocCtrlEventRegister(std::span<const u8> input, std::vecto
    return NvResult::Success;
 }
-NvResult nvhost_ctrl::IocCtrlEventUnregister(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::IocCtrlEventUnregister(std::span<const u8> input, std::span<u8> output) {
    IocCtrlEventUnregisterParams params{};
    std::memcpy(&params, input.data(), sizeof(params));
    const u32 event_id = params.user_event_id & 0x00FF;
@@ -262,8 +262,7 @@ NvResult nvhost_ctrl::IocCtrlEventUnregister(std::span<const u8> input, std::vec
    return FreeEvent(event_id);
 }
-NvResult nvhost_ctrl::IocCtrlEventUnregisterBatch(std::span<const u8> input,
+NvResult nvhost_ctrl::IocCtrlEventUnregisterBatch(std::span<const u8> input, std::span<u8> output) {
-                                                  std::vector<u8>& output) {
    IocCtrlEventUnregisterBatchParams params{};
    std::memcpy(&params, input.data(), sizeof(params));
    u64 event_mask = params.user_events;
@@ -281,7 +280,7 @@ NvResult nvhost_ctrl::IocCtrlEventUnregisterBatch(std::span<const u8> input,
    return NvResult::Success;
 }
-NvResult nvhost_ctrl::IocCtrlClearEventWait(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::IocCtrlClearEventWait(std::span<const u8> input, std::span<u8> output) {
    IocCtrlEventClearParams params{};
    std::memcpy(&params, input.data(), sizeof(params));
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl.h b/src/core/hle/service/nvdrv/devices/nvhost_ctrl.h
index dd2e7888a..2efed4862 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl.h
@@ -26,11 +26,11 @@ public:
    ~nvhost_ctrl() override;
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
@@ -186,13 +186,12 @@ private:
    static_assert(sizeof(IocCtrlEventUnregisterBatchParams) == 8,
                  "IocCtrlEventKill is incorrect size");
-    NvResult NvOsGetConfigU32(std::span<const u8> input, std::vector<u8>& output);
+    NvResult NvOsGetConfigU32(std::span<const u8> input, std::span<u8> output);
-    NvResult IocCtrlEventWait(std::span<const u8> input, std::vector<u8>& output,
+    NvResult IocCtrlEventWait(std::span<const u8> input, std::span<u8> output, bool is_allocation);
-                              bool is_allocation);
+    NvResult IocCtrlEventRegister(std::span<const u8> input, std::span<u8> output);
-    NvResult IocCtrlEventRegister(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocCtrlEventUnregister(std::span<const u8> input, std::span<u8> output);
-    NvResult IocCtrlEventUnregister(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocCtrlEventUnregisterBatch(std::span<const u8> input, std::span<u8> output);
-    NvResult IocCtrlEventUnregisterBatch(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocCtrlClearEventWait(std::span<const u8> input, std::span<u8> output);
-    NvResult IocCtrlClearEventWait(std::span<const u8> input, std::vector<u8>& output);
    NvResult FreeEvent(u32 slot);
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp
index be3c083db..6081d92e9 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp
@@ -22,7 +22,7 @@ nvhost_ctrl_gpu::~nvhost_ctrl_gpu() {
 }
 NvResult nvhost_ctrl_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                                 std::vector<u8>& output) {
+                                 std::span<u8> output) {
    switch (command.group) {
    case 'G':
        switch (command.cmd) {
@@ -54,13 +54,13 @@ NvResult nvhost_ctrl_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8>
 }
 NvResult nvhost_ctrl_gpu::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                                 std::span<const u8> inline_input, std::vector<u8>& output) {
+                                 std::span<const u8> inline_input, std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
 NvResult nvhost_ctrl_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                                 std::vector<u8>& output, std::vector<u8>& inline_output) {
+                                 std::span<u8> output, std::span<u8> inline_output) {
    switch (command.group) {
    case 'G':
        switch (command.cmd) {
@@ -82,7 +82,7 @@ NvResult nvhost_ctrl_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8>
 void nvhost_ctrl_gpu::OnOpen(DeviceFD fd) {}
 void nvhost_ctrl_gpu::OnClose(DeviceFD fd) {}
-NvResult nvhost_ctrl_gpu::GetCharacteristics(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::GetCharacteristics(std::span<const u8> input, std::span<u8> output) {
    LOG_DEBUG(Service_NVDRV, "called");
    IoctlCharacteristics params{};
    std::memcpy(&params, input.data(), input.size());
@@ -127,8 +127,8 @@ NvResult nvhost_ctrl_gpu::GetCharacteristics(std::span<const u8> input, std::vec
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::GetCharacteristics(std::span<const u8> input, std::vector<u8>& output,
+NvResult nvhost_ctrl_gpu::GetCharacteristics(std::span<const u8> input, std::span<u8> output,
-                                             std::vector<u8>& inline_output) {
+                                             std::span<u8> inline_output) {
    LOG_DEBUG(Service_NVDRV, "called");
    IoctlCharacteristics params{};
    std::memcpy(&params, input.data(), input.size());
@@ -175,7 +175,7 @@ NvResult nvhost_ctrl_gpu::GetCharacteristics(std::span<const u8> input, std::vec
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::GetTPCMasks(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::GetTPCMasks(std::span<const u8> input, std::span<u8> output) {
    IoctlGpuGetTpcMasksArgs params{};
    std::memcpy(&params, input.data(), input.size());
    LOG_DEBUG(Service_NVDRV, "called, mask_buffer_size=0x{:X}", params.mask_buffer_size);
@@ -186,8 +186,8 @@ NvResult nvhost_ctrl_gpu::GetTPCMasks(std::span<const u8> input, std::vector<u8>
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::GetTPCMasks(std::span<const u8> input, std::vector<u8>& output,
+NvResult nvhost_ctrl_gpu::GetTPCMasks(std::span<const u8> input, std::span<u8> output,
-                                      std::vector<u8>& inline_output) {
+                                      std::span<u8> inline_output) {
    IoctlGpuGetTpcMasksArgs params{};
    std::memcpy(&params, input.data(), input.size());
    LOG_DEBUG(Service_NVDRV, "called, mask_buffer_size=0x{:X}", params.mask_buffer_size);
@@ -199,7 +199,7 @@ NvResult nvhost_ctrl_gpu::GetTPCMasks(std::span<const u8> input, std::vector<u8>
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::GetActiveSlotMask(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::GetActiveSlotMask(std::span<const u8> input, std::span<u8> output) {
    LOG_DEBUG(Service_NVDRV, "called");
    IoctlActiveSlotMask params{};
@@ -212,7 +212,7 @@ NvResult nvhost_ctrl_gpu::GetActiveSlotMask(std::span<const u8> input, std::vect
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::ZCullGetCtxSize(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::ZCullGetCtxSize(std::span<const u8> input, std::span<u8> output) {
    LOG_DEBUG(Service_NVDRV, "called");
    IoctlZcullGetCtxSize params{};
@@ -224,7 +224,7 @@ NvResult nvhost_ctrl_gpu::ZCullGetCtxSize(std::span<const u8> input, std::vector
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::ZCullGetInfo(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::ZCullGetInfo(std::span<const u8> input, std::span<u8> output) {
    LOG_DEBUG(Service_NVDRV, "called");
    IoctlNvgpuGpuZcullGetInfoArgs params{};
@@ -247,7 +247,7 @@ NvResult nvhost_ctrl_gpu::ZCullGetInfo(std::span<const u8> input, std::vector<u8
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::ZBCSetTable(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::ZBCSetTable(std::span<const u8> input, std::span<u8> output) {
    LOG_WARNING(Service_NVDRV, "(STUBBED) called");
    IoctlZbcSetTable params{};
@@ -263,7 +263,7 @@ NvResult nvhost_ctrl_gpu::ZBCSetTable(std::span<const u8> input, std::vector<u8>
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::ZBCQueryTable(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::ZBCQueryTable(std::span<const u8> input, std::span<u8> output) {
    LOG_WARNING(Service_NVDRV, "(STUBBED) called");
    IoctlZbcQueryTable params{};
@@ -273,7 +273,7 @@ NvResult nvhost_ctrl_gpu::ZBCQueryTable(std::span<const u8> input, std::vector<u
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::FlushL2(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::FlushL2(std::span<const u8> input, std::span<u8> output) {
    LOG_WARNING(Service_NVDRV, "(STUBBED) called");
    IoctlFlushL2 params{};
@@ -283,7 +283,7 @@ NvResult nvhost_ctrl_gpu::FlushL2(std::span<const u8> input, std::vector<u8>& ou
    return NvResult::Success;
 }
-NvResult nvhost_ctrl_gpu::GetGpuTime(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::GetGpuTime(std::span<const u8> input, std::span<u8> output) {
    LOG_DEBUG(Service_NVDRV, "called");
    IoctlGetGpuTime params{};
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h
index b9333d9d3..97995551c 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h
@@ -22,11 +22,11 @@ public:
    ~nvhost_ctrl_gpu() override;
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
@@ -151,21 +151,21 @@ private:
    };
    static_assert(sizeof(IoctlGetGpuTime) == 0x10, "IoctlGetGpuTime is incorrect size");
-    NvResult GetCharacteristics(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetCharacteristics(std::span<const u8> input, std::span<u8> output);
-    NvResult GetCharacteristics(std::span<const u8> input, std::vector<u8>& output,
+    NvResult GetCharacteristics(std::span<const u8> input, std::span<u8> output,
-                                std::vector<u8>& inline_output);
+                                std::span<u8> inline_output);
-    NvResult GetTPCMasks(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetTPCMasks(std::span<const u8> input, std::span<u8> output);
-    NvResult GetTPCMasks(std::span<const u8> input, std::vector<u8>& output,
+    NvResult GetTPCMasks(std::span<const u8> input, std::span<u8> output,
-                         std::vector<u8>& inline_output);
+                         std::span<u8> inline_output);
-    NvResult GetActiveSlotMask(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetActiveSlotMask(std::span<const u8> input, std::span<u8> output);
-    NvResult ZCullGetCtxSize(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZCullGetCtxSize(std::span<const u8> input, std::span<u8> output);
-    NvResult ZCullGetInfo(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZCullGetInfo(std::span<const u8> input, std::span<u8> output);
-    NvResult ZBCSetTable(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZBCSetTable(std::span<const u8> input, std::span<u8> output);
-    NvResult ZBCQueryTable(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZBCQueryTable(std::span<const u8> input, std::span<u8> output);
-    NvResult FlushL2(std::span<const u8> input, std::vector<u8>& output);
+    NvResult FlushL2(std::span<const u8> input, std::span<u8> output);
-    NvResult GetGpuTime(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetGpuTime(std::span<const u8> input, std::span<u8> output);
    EventInterface& events_interface;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp b/src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp
index 453a965dc..46a25fcab 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp
@@ -47,7 +47,7 @@ nvhost_gpu::~nvhost_gpu() {
 }
 NvResult nvhost_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::vector<u8>& output) {
+                            std::span<u8> output) {
    switch (command.group) {
    case 0x0:
        switch (command.cmd) {
@@ -99,7 +99,7 @@ NvResult nvhost_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> inpu
 };
 NvResult nvhost_gpu::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::span<const u8> inline_input, std::vector<u8>& output) {
+                            std::span<const u8> inline_input, std::span<u8> output) {
    switch (command.group) {
    case 'H':
        switch (command.cmd) {
@@ -113,7 +113,7 @@ NvResult nvhost_gpu::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> inpu
 }
 NvResult nvhost_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::vector<u8>& output, std::vector<u8>& inline_output) {
+                            std::span<u8> output, std::span<u8> inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
@@ -121,7 +121,7 @@ NvResult nvhost_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> inpu
 void nvhost_gpu::OnOpen(DeviceFD fd) {}
 void nvhost_gpu::OnClose(DeviceFD fd) {}
-NvResult nvhost_gpu::SetNVMAPfd(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::SetNVMAPfd(std::span<const u8> input, std::span<u8> output) {
    IoctlSetNvmapFD params{};
    std::memcpy(&params, input.data(), input.size());
    LOG_DEBUG(Service_NVDRV, "called, fd={}", params.nvmap_fd);
@@ -130,7 +130,7 @@ NvResult nvhost_gpu::SetNVMAPfd(std::span<const u8> input, std::vector<u8>& outp
    return NvResult::Success;
 }
-NvResult nvhost_gpu::SetClientData(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::SetClientData(std::span<const u8> input, std::span<u8> output) {
    LOG_DEBUG(Service_NVDRV, "called");
    IoctlClientData params{};
@@ -139,7 +139,7 @@ NvResult nvhost_gpu::SetClientData(std::span<const u8> input, std::vector<u8>& o
    return NvResult::Success;
 }
-NvResult nvhost_gpu::GetClientData(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::GetClientData(std::span<const u8> input, std::span<u8> output) {
    LOG_DEBUG(Service_NVDRV, "called");
    IoctlClientData params{};
@@ -149,7 +149,7 @@ NvResult nvhost_gpu::GetClientData(std::span<const u8> input, std::vector<u8>& o
    return NvResult::Success;
 }
-NvResult nvhost_gpu::ZCullBind(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::ZCullBind(std::span<const u8> input, std::span<u8> output) {
    std::memcpy(&zcull_params, input.data(), input.size());
    LOG_DEBUG(Service_NVDRV, "called, gpu_va={:X}, mode={:X}", zcull_params.gpu_va,
              zcull_params.mode);
@@ -158,7 +158,7 @@ NvResult nvhost_gpu::ZCullBind(std::span<const u8> input, std::vector<u8>& outpu
    return NvResult::Success;
 }
-NvResult nvhost_gpu::SetErrorNotifier(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::SetErrorNotifier(std::span<const u8> input, std::span<u8> output) {
    IoctlSetErrorNotifier params{};
    std::memcpy(&params, input.data(), input.size());
    LOG_WARNING(Service_NVDRV, "(STUBBED) called, offset={:X}, size={:X}, mem={:X}", params.offset,
@@ -168,14 +168,14 @@ NvResult nvhost_gpu::SetErrorNotifier(std::span<const u8> input, std::vector<u8>
    return NvResult::Success;
 }
-NvResult nvhost_gpu::SetChannelPriority(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::SetChannelPriority(std::span<const u8> input, std::span<u8> output) {
    std::memcpy(&channel_priority, input.data(), input.size());
    LOG_DEBUG(Service_NVDRV, "(STUBBED) called, priority={:X}", channel_priority);
    return NvResult::Success;
 }
-NvResult nvhost_gpu::AllocGPFIFOEx2(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::AllocGPFIFOEx2(std::span<const u8> input, std::span<u8> output) {
    IoctlAllocGpfifoEx2 params{};
    std::memcpy(&params, input.data(), input.size());
    LOG_WARNING(Service_NVDRV,
@@ -197,7 +197,7 @@ NvResult nvhost_gpu::AllocGPFIFOEx2(std::span<const u8> input, std::vector<u8>&
    return NvResult::Success;
 }
-NvResult nvhost_gpu::AllocateObjectContext(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::AllocateObjectContext(std::span<const u8> input, std::span<u8> output) {
    IoctlAllocObjCtx params{};
    std::memcpy(&params, input.data(), input.size());
    LOG_WARNING(Service_NVDRV, "(STUBBED) called, class_num={:X}, flags={:X}", params.class_num,
@@ -208,7 +208,8 @@ NvResult nvhost_gpu::AllocateObjectContext(std::span<const u8> input, std::vecto
    return NvResult::Success;
 }
-static std::vector<Tegra::CommandHeader> BuildWaitCommandList(NvFence fence) {
+static boost::container::small_vector<Tegra::CommandHeader, 512> BuildWaitCommandList(
+    NvFence fence) {
    return {
        Tegra::BuildCommandHeader(Tegra::BufferMethods::SyncpointPayload, 1,
                                  Tegra::SubmissionMode::Increasing),
@@ -219,35 +220,35 @@ static std::vector<Tegra::CommandHeader> BuildWaitCommandList(NvFence fence) {
    };
 }
-static std::vector<Tegra::CommandHeader> BuildIncrementCommandList(NvFence fence) {
+static boost::container::small_vector<Tegra::CommandHeader, 512> BuildIncrementCommandList(
-    std::vector<Tegra::CommandHeader> result{
+    NvFence fence) {
+    boost::container::small_vector<Tegra::CommandHeader, 512> result{
        Tegra::BuildCommandHeader(Tegra::BufferMethods::SyncpointPayload, 1,
                                  Tegra::SubmissionMode::Increasing),
        {}};
    for (u32 count = 0; count < 2; ++count) {
-        result.emplace_back(Tegra::BuildCommandHeader(Tegra::BufferMethods::SyncpointOperation, 1,
+        result.push_back(Tegra::BuildCommandHeader(Tegra::BufferMethods::SyncpointOperation, 1,
-                                                      Tegra::SubmissionMode::Increasing));
+                                                   Tegra::SubmissionMode::Increasing));
-        result.emplace_back(
+        result.push_back(
            BuildFenceAction(Tegra::Engines::Puller::FenceOperation::Increment, fence.id));
    }
    return result;
 }
-static std::vector<Tegra::CommandHeader> BuildIncrementWithWfiCommandList(NvFence fence) {
+static boost::container::small_vector<Tegra::CommandHeader, 512> BuildIncrementWithWfiCommandList(
-    std::vector<Tegra::CommandHeader> result{
+    NvFence fence) {
+    boost::container::small_vector<Tegra::CommandHeader, 512> result{
        Tegra::BuildCommandHeader(Tegra::BufferMethods::WaitForIdle, 1,
                                  Tegra::SubmissionMode::Increasing),
        {}};
-    const std::vector<Tegra::CommandHeader> increment{BuildIncrementCommandList(fence)};
+    auto increment_list{BuildIncrementCommandList(fence)};
+    result.insert(result.end(), increment_list.begin(), increment_list.end());
-    result.insert(result.end(), increment.begin(), increment.end());
    return result;
 }
-NvResult nvhost_gpu::SubmitGPFIFOImpl(IoctlSubmitGpfifo& params, std::vector<u8>& output,
+NvResult nvhost_gpu::SubmitGPFIFOImpl(IoctlSubmitGpfifo& params, std::span<u8> output,
                                      Tegra::CommandList&& entries) {
    LOG_TRACE(Service_NVDRV, "called, gpfifo={:X}, num_entries={:X}, flags={:X}", params.address,
              params.num_entries, params.flags.raw);
@@ -293,7 +294,7 @@ NvResult nvhost_gpu::SubmitGPFIFOImpl(IoctlSubmitGpfifo& params, std::vector<u8>
    return NvResult::Success;
 }
-NvResult nvhost_gpu::SubmitGPFIFOBase(std::span<const u8> input, std::vector<u8>& output,
+NvResult nvhost_gpu::SubmitGPFIFOBase(std::span<const u8> input, std::span<u8> output,
                                      bool kickoff) {
    if (input.size() < sizeof(IoctlSubmitGpfifo)) {
        UNIMPLEMENTED();
@@ -315,7 +316,7 @@ NvResult nvhost_gpu::SubmitGPFIFOBase(std::span<const u8> input, std::vector<u8>
 }
 NvResult nvhost_gpu::SubmitGPFIFOBase(std::span<const u8> input, std::span<const u8> input_inline,
-                                      std::vector<u8>& output) {
+                                      std::span<u8> output) {
    if (input.size() < sizeof(IoctlSubmitGpfifo)) {
        UNIMPLEMENTED();
        return NvResult::InvalidSize;
@@ -327,7 +328,7 @@ NvResult nvhost_gpu::SubmitGPFIFOBase(std::span<const u8> input, std::span<const
    return SubmitGPFIFOImpl(params, output, std::move(entries));
 }
-NvResult nvhost_gpu::GetWaitbase(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::GetWaitbase(std::span<const u8> input, std::span<u8> output) {
    IoctlGetWaitbase params{};
    std::memcpy(&params, input.data(), sizeof(IoctlGetWaitbase));
    LOG_INFO(Service_NVDRV, "called, unknown=0x{:X}", params.unknown);
@@ -337,7 +338,7 @@ NvResult nvhost_gpu::GetWaitbase(std::span<const u8> input, std::vector<u8>& out
    return NvResult::Success;
 }
-NvResult nvhost_gpu::ChannelSetTimeout(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::ChannelSetTimeout(std::span<const u8> input, std::span<u8> output) {
    IoctlChannelSetTimeout params{};
    std::memcpy(&params, input.data(), sizeof(IoctlChannelSetTimeout));
    LOG_INFO(Service_NVDRV, "called, timeout=0x{:X}", params.timeout);
@@ -345,7 +346,7 @@ NvResult nvhost_gpu::ChannelSetTimeout(std::span<const u8> input, std::vector<u8
    return NvResult::Success;
 }
-NvResult nvhost_gpu::ChannelSetTimeslice(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_gpu::ChannelSetTimeslice(std::span<const u8> input, std::span<u8> output) {
    IoctlSetTimeslice params{};
    std::memcpy(&params, input.data(), sizeof(IoctlSetTimeslice));
    LOG_INFO(Service_NVDRV, "called, timeslice=0x{:X}", params.timeslice);
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_gpu.h b/src/core/hle/service/nvdrv/devices/nvhost_gpu.h
index 3ca58202d..529c20526 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_gpu.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_gpu.h
@@ -41,11 +41,11 @@ public:
    ~nvhost_gpu() override;
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
@@ -186,23 +186,23 @@ private:
    u32_le channel_priority{};
    u32_le channel_timeslice{};
-    NvResult SetNVMAPfd(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetNVMAPfd(std::span<const u8> input, std::span<u8> output);
-    NvResult SetClientData(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetClientData(std::span<const u8> input, std::span<u8> output);
-    NvResult GetClientData(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetClientData(std::span<const u8> input, std::span<u8> output);
-    NvResult ZCullBind(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZCullBind(std::span<const u8> input, std::span<u8> output);
-    NvResult SetErrorNotifier(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetErrorNotifier(std::span<const u8> input, std::span<u8> output);
-    NvResult SetChannelPriority(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetChannelPriority(std::span<const u8> input, std::span<u8> output);
-    NvResult AllocGPFIFOEx2(std::span<const u8> input, std::vector<u8>& output);
+    NvResult AllocGPFIFOEx2(std::span<const u8> input, std::span<u8> output);
-    NvResult AllocateObjectContext(std::span<const u8> input, std::vector<u8>& output);
+    NvResult AllocateObjectContext(std::span<const u8> input, std::span<u8> output);
-    NvResult SubmitGPFIFOImpl(IoctlSubmitGpfifo& params, std::vector<u8>& output,
+    NvResult SubmitGPFIFOImpl(IoctlSubmitGpfifo& params, std::span<u8> output,
                              Tegra::CommandList&& entries);
-    NvResult SubmitGPFIFOBase(std::span<const u8> input, std::vector<u8>& output,
+    NvResult SubmitGPFIFOBase(std::span<const u8> input, std::span<u8> output,
                              bool kickoff = false);
    NvResult SubmitGPFIFOBase(std::span<const u8> input, std::span<const u8> input_inline,
-                              std::vector<u8>& output);
+                              std::span<u8> output);
-    NvResult GetWaitbase(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetWaitbase(std::span<const u8> input, std::span<u8> output);
-    NvResult ChannelSetTimeout(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ChannelSetTimeout(std::span<const u8> input, std::span<u8> output);
-    NvResult ChannelSetTimeslice(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ChannelSetTimeslice(std::span<const u8> input, std::span<u8> output);
    EventInterface& events_interface;
    NvCore::Container& core;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
index dc45169ad..a174442a6 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
@@ -16,7 +16,7 @@ nvhost_nvdec::nvhost_nvdec(Core::System& system_, NvCore::Container& core_)
 nvhost_nvdec::~nvhost_nvdec() = default;
 NvResult nvhost_nvdec::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::vector<u8>& output) {
+                              std::span<u8> output) {
    switch (command.group) {
    case 0x0:
        switch (command.cmd) {
@@ -56,13 +56,13 @@ NvResult nvhost_nvdec::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> in
 }
 NvResult nvhost_nvdec::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::span<const u8> inline_input, std::vector<u8>& output) {
+                              std::span<const u8> inline_input, std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
 NvResult nvhost_nvdec::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::vector<u8>& output, std::vector<u8>& inline_output) {
+                              std::span<u8> output, std::span<u8> inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
index 0d615bbcb..ad2233c49 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
@@ -14,11 +14,11 @@ public:
    ~nvhost_nvdec() override;
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.cpp b/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.cpp
index 1ab51f10b..61649aa4a 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.cpp
@@ -36,7 +36,7 @@ std::size_t SliceVectors(std::span<const u8> input, std::vector<T>& dst, std::si
 // Writes the data in src to an offset into the dst vector. The offset is specified in bytes
 // Returns the number of bytes written into dst.
 template <typename T>
-std::size_t WriteVectors(std::vector<u8>& dst, const std::vector<T>& src, std::size_t offset) {
+std::size_t WriteVectors(std::span<u8> dst, const std::vector<T>& src, std::size_t offset) {
    if (src.empty()) {
        return 0;
    }
@@ -72,8 +72,7 @@ NvResult nvhost_nvdec_common::SetNVMAPfd(std::span<const u8> input) {
    return NvResult::Success;
 }
-NvResult nvhost_nvdec_common::Submit(DeviceFD fd, std::span<const u8> input,
+NvResult nvhost_nvdec_common::Submit(DeviceFD fd, std::span<const u8> input, std::span<u8> output) {
-                                     std::vector<u8>& output) {
    IoctlSubmit params{};
    std::memcpy(&params, input.data(), sizeof(IoctlSubmit));
    LOG_DEBUG(Service_NVDRV, "called NVDEC Submit, cmd_buffer_count={}", params.cmd_buffer_count);
@@ -121,7 +120,7 @@ NvResult nvhost_nvdec_common::Submit(DeviceFD fd, std::span<const u8> input,
    return NvResult::Success;
 }
-NvResult nvhost_nvdec_common::GetSyncpoint(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::GetSyncpoint(std::span<const u8> input, std::span<u8> output) {
    IoctlGetSyncpoint params{};
    std::memcpy(&params, input.data(), sizeof(IoctlGetSyncpoint));
    LOG_DEBUG(Service_NVDRV, "called GetSyncpoint, id={}", params.param);
@@ -133,7 +132,7 @@ NvResult nvhost_nvdec_common::GetSyncpoint(std::span<const u8> input, std::vecto
    return NvResult::Success;
 }
-NvResult nvhost_nvdec_common::GetWaitbase(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::GetWaitbase(std::span<const u8> input, std::span<u8> output) {
    IoctlGetWaitbase params{};
    LOG_CRITICAL(Service_NVDRV, "called WAITBASE");
    std::memcpy(&params, input.data(), sizeof(IoctlGetWaitbase));
@@ -142,7 +141,7 @@ NvResult nvhost_nvdec_common::GetWaitbase(std::span<const u8> input, std::vector
    return NvResult::Success;
 }
-NvResult nvhost_nvdec_common::MapBuffer(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::MapBuffer(std::span<const u8> input, std::span<u8> output) {
    IoctlMapBuffer params{};
    std::memcpy(&params, input.data(), sizeof(IoctlMapBuffer));
    std::vector<MapBufferEntry> cmd_buffer_handles(params.num_entries);
@@ -159,7 +158,7 @@ NvResult nvhost_nvdec_common::MapBuffer(std::span<const u8> input, std::vector<u
    return NvResult::Success;
 }
-NvResult nvhost_nvdec_common::UnmapBuffer(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::UnmapBuffer(std::span<const u8> input, std::span<u8> output) {
    IoctlMapBuffer params{};
    std::memcpy(&params, input.data(), sizeof(IoctlMapBuffer));
    std::vector<MapBufferEntry> cmd_buffer_handles(params.num_entries);
@@ -173,7 +172,7 @@ NvResult nvhost_nvdec_common::UnmapBuffer(std::span<const u8> input, std::vector
    return NvResult::Success;
 }
-NvResult nvhost_nvdec_common::SetSubmitTimeout(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::SetSubmitTimeout(std::span<const u8> input, std::span<u8> output) {
    std::memcpy(&submit_timeout, input.data(), input.size());
    LOG_WARNING(Service_NVDRV, "(STUBBED) called");
    return NvResult::Success;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.h b/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.h
index 5af26a26f..9bb573bfe 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.h
@@ -108,12 +108,12 @@ protected:
    /// Ioctl command implementations
    NvResult SetNVMAPfd(std::span<const u8> input);
-    NvResult Submit(DeviceFD fd, std::span<const u8> input, std::vector<u8>& output);
+    NvResult Submit(DeviceFD fd, std::span<const u8> input, std::span<u8> output);
-    NvResult GetSyncpoint(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetSyncpoint(std::span<const u8> input, std::span<u8> output);
-    NvResult GetWaitbase(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetWaitbase(std::span<const u8> input, std::span<u8> output);
-    NvResult MapBuffer(std::span<const u8> input, std::vector<u8>& output);
+    NvResult MapBuffer(std::span<const u8> input, std::span<u8> output);
-    NvResult UnmapBuffer(std::span<const u8> input, std::vector<u8>& output);
+    NvResult UnmapBuffer(std::span<const u8> input, std::span<u8> output);
-    NvResult SetSubmitTimeout(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetSubmitTimeout(std::span<const u8> input, std::span<u8> output);
    Kernel::KEvent* QueryEvent(u32 event_id) override;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.cpp b/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.cpp
index 39f30e7c8..a05c8cdae 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.cpp
@@ -13,7 +13,7 @@ nvhost_nvjpg::nvhost_nvjpg(Core::System& system_) : nvdevice{system_} {}
 nvhost_nvjpg::~nvhost_nvjpg() = default;
 NvResult nvhost_nvjpg::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::vector<u8>& output) {
+                              std::span<u8> output) {
    switch (command.group) {
    case 'H':
        switch (command.cmd) {
@@ -32,13 +32,13 @@ NvResult nvhost_nvjpg::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> in
 }
 NvResult nvhost_nvjpg::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::span<const u8> inline_input, std::vector<u8>& output) {
+                              std::span<const u8> inline_input, std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
 NvResult nvhost_nvjpg::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                              std::vector<u8>& output, std::vector<u8>& inline_output) {
+                              std::span<u8> output, std::span<u8> inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
@@ -46,7 +46,7 @@ NvResult nvhost_nvjpg::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> in
 void nvhost_nvjpg::OnOpen(DeviceFD fd) {}
 void nvhost_nvjpg::OnClose(DeviceFD fd) {}
-NvResult nvhost_nvjpg::SetNVMAPfd(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvhost_nvjpg::SetNVMAPfd(std::span<const u8> input, std::span<u8> output) {
    IoctlSetNvmapFD params{};
    std::memcpy(&params, input.data(), input.size());
    LOG_DEBUG(Service_NVDRV, "called, fd={}", params.nvmap_fd);
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.h b/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.h
index 41b57e872..5623e0d47 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.h
@@ -16,11 +16,11 @@ public:
    ~nvhost_nvjpg() override;
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
@@ -33,7 +33,7 @@ private:
    s32_le nvmap_fd{};
-    NvResult SetNVMAPfd(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetNVMAPfd(std::span<const u8> input, std::span<u8> output);
 };
 } // namespace Service::Nvidia::Devices
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp b/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
index b0ea402a7..c0b8684c3 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
@@ -16,7 +16,7 @@ nvhost_vic::nvhost_vic(Core::System& system_, NvCore::Container& core_)
 nvhost_vic::~nvhost_vic() = default;
 NvResult nvhost_vic::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::vector<u8>& output) {
+                            std::span<u8> output) {
    switch (command.group) {
    case 0x0:
        switch (command.cmd) {
@@ -56,13 +56,13 @@ NvResult nvhost_vic::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> inpu
 }
 NvResult nvhost_vic::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::span<const u8> inline_input, std::vector<u8>& output) {
+                            std::span<const u8> inline_input, std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
 NvResult nvhost_vic::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                            std::vector<u8>& output, std::vector<u8>& inline_output) {
+                            std::span<u8> output, std::span<u8> inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_vic.h b/src/core/hle/service/nvdrv/devices/nvhost_vic.h
index b5e350a83..cadbcb0a5 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_vic.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_vic.h
@@ -13,11 +13,11 @@ public:
    ~nvhost_vic();
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
diff --git a/src/core/hle/service/nvdrv/devices/nvmap.cpp b/src/core/hle/service/nvdrv/devices/nvmap.cpp
index 07417f045..e7f7e273b 100644
--- a/src/core/hle/service/nvdrv/devices/nvmap.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvmap.cpp
@@ -26,7 +26,7 @@ nvmap::nvmap(Core::System& system_, NvCore::Container& container_)
 nvmap::~nvmap() = default;
 NvResult nvmap::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                       std::vector<u8>& output) {
+                       std::span<u8> output) {
    switch (command.group) {
    case 0x1:
        switch (command.cmd) {
@@ -55,13 +55,13 @@ NvResult nvmap::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
 }
 NvResult nvmap::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                       std::span<const u8> inline_input, std::vector<u8>& output) {
+                       std::span<const u8> inline_input, std::span<u8> output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
-NvResult nvmap::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
+NvResult nvmap::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                       std::vector<u8>& output, std::vector<u8>& inline_output) {
+                       std::span<u8> inline_output) {
    UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
    return NvResult::NotImplemented;
 }
@@ -69,7 +69,7 @@ NvResult nvmap::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
 void nvmap::OnOpen(DeviceFD fd) {}
 void nvmap::OnClose(DeviceFD fd) {}
-NvResult nvmap::IocCreate(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvmap::IocCreate(std::span<const u8> input, std::span<u8> output) {
    IocCreateParams params;
    std::memcpy(&params, input.data(), sizeof(params));
    LOG_DEBUG(Service_NVDRV, "called, size=0x{:08X}", params.size);
@@ -89,7 +89,7 @@ NvResult nvmap::IocCreate(std::span<const u8> input, std::vector<u8>& output) {
    return NvResult::Success;
 }
-NvResult nvmap::IocAlloc(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvmap::IocAlloc(std::span<const u8> input, std::span<u8> output) {
    IocAllocParams params;
    std::memcpy(&params, input.data(), sizeof(params));
    LOG_DEBUG(Service_NVDRV, "called, addr={:X}", params.address);
@@ -137,7 +137,7 @@ NvResult nvmap::IocAlloc(std::span<const u8> input, std::vector<u8>& output) {
    return result;
 }
-NvResult nvmap::IocGetId(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvmap::IocGetId(std::span<const u8> input, std::span<u8> output) {
    IocGetIdParams params;
    std::memcpy(&params, input.data(), sizeof(params));
@@ -161,7 +161,7 @@ NvResult nvmap::IocGetId(std::span<const u8> input, std::vector<u8>& output) {
    return NvResult::Success;
 }
-NvResult nvmap::IocFromId(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvmap::IocFromId(std::span<const u8> input, std::span<u8> output) {
    IocFromIdParams params;
    std::memcpy(&params, input.data(), sizeof(params));
@@ -192,7 +192,7 @@ NvResult nvmap::IocFromId(std::span<const u8> input, std::vector<u8>& output) {
    return NvResult::Success;
 }
-NvResult nvmap::IocParam(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvmap::IocParam(std::span<const u8> input, std::span<u8> output) {
    enum class ParamTypes { Size = 1, Alignment = 2, Base = 3, Heap = 4, Kind = 5, Compr = 6 };
    IocParamParams params;
@@ -241,7 +241,7 @@ NvResult nvmap::IocParam(std::span<const u8> input, std::vector<u8>& output) {
    return NvResult::Success;
 }
-NvResult nvmap::IocFree(std::span<const u8> input, std::vector<u8>& output) {
+NvResult nvmap::IocFree(std::span<const u8> input, std::span<u8> output) {
    IocFreeParams params;
    std::memcpy(&params, input.data(), sizeof(params));
diff --git a/src/core/hle/service/nvdrv/devices/nvmap.h b/src/core/hle/service/nvdrv/devices/nvmap.h
index 82bd3b118..40c65b430 100644
--- a/src/core/hle/service/nvdrv/devices/nvmap.h
+++ b/src/core/hle/service/nvdrv/devices/nvmap.h
@@ -27,11 +27,11 @@ public:
    nvmap& operator=(const nvmap&) = delete;
    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::vector<u8>& output) override;
+                    std::span<u8> output) override;
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+                    std::span<const u8> inline_input, std::span<u8> output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output) override;
+                    std::span<u8> inline_output) override;
    void OnOpen(DeviceFD fd) override;
    void OnClose(DeviceFD fd) override;
@@ -106,12 +106,12 @@ private:
    };
    static_assert(sizeof(IocGetIdParams) == 8, "IocGetIdParams has wrong size");
-    NvResult IocCreate(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocCreate(std::span<const u8> input, std::span<u8> output);
-    NvResult IocAlloc(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocAlloc(std::span<const u8> input, std::span<u8> output);
-    NvResult IocGetId(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocGetId(std::span<const u8> input, std::span<u8> output);
-    NvResult IocFromId(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocFromId(std::span<const u8> input, std::span<u8> output);
-    NvResult IocParam(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocParam(std::span<const u8> input, std::span<u8> output);
-    NvResult IocFree(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocFree(std::span<const u8> input, std::span<u8> output);
    NvCore::Container& container;
    NvCore::NvMap& file;
diff --git a/src/core/hle/service/nvdrv/nvdrv.cpp b/src/core/hle/service/nvdrv/nvdrv.cpp
index 3d774eec4..9e46ee8dd 100644
--- a/src/core/hle/service/nvdrv/nvdrv.cpp
+++ b/src/core/hle/service/nvdrv/nvdrv.cpp
@@ -130,7 +130,7 @@ DeviceFD Module::Open(const std::string& device_name) {
 }
 NvResult Module::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                        std::vector<u8>& output) {
+                        std::span<u8> output) {
    if (fd < 0) {
        LOG_ERROR(Service_NVDRV, "Invalid DeviceFD={}!", fd);
        return NvResult::InvalidState;
@@ -147,7 +147,7 @@ NvResult Module::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
 }
 NvResult Module::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                        std::span<const u8> inline_input, std::vector<u8>& output) {
+                        std::span<const u8> inline_input, std::span<u8> output) {
    if (fd < 0) {
        LOG_ERROR(Service_NVDRV, "Invalid DeviceFD={}!", fd);
        return NvResult::InvalidState;
@@ -163,8 +163,8 @@ NvResult Module::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
    return itr->second->Ioctl2(fd, command, input, inline_input, output);
 }
-NvResult Module::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
+NvResult Module::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                        std::vector<u8>& output, std::vector<u8>& inline_output) {
+                        std::span<u8> inline_output) {
    if (fd < 0) {
        LOG_ERROR(Service_NVDRV, "Invalid DeviceFD={}!", fd);
        return NvResult::InvalidState;
diff --git a/src/core/hle/service/nvdrv/nvdrv.h b/src/core/hle/service/nvdrv/nvdrv.h
index 668be742b..d8622b3ca 100644
--- a/src/core/hle/service/nvdrv/nvdrv.h
+++ b/src/core/hle/service/nvdrv/nvdrv.h
@@ -80,13 +80,13 @@ public:
    DeviceFD Open(const std::string& device_name);
    /// Sends an ioctl command to the specified file descriptor.
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output);
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output);
    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
-                    std::span<const u8> inline_input, std::vector<u8>& output);
+                    std::span<const u8> inline_input, std::span<u8> output);
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::span<u8> output,
-                    std::vector<u8>& inline_output);
+                    std::span<u8> inline_output);
    /// Closes a device file descriptor and returns operation success.
    NvResult Close(DeviceFD fd);
diff --git a/src/core/hle/service/nvdrv/nvdrv_interface.cpp b/src/core/hle/service/nvdrv/nvdrv_interface.cpp
index d010a1e03..348207e25 100644
--- a/src/core/hle/service/nvdrv/nvdrv_interface.cpp
+++ b/src/core/hle/service/nvdrv/nvdrv_interface.cpp
@@ -63,12 +63,12 @@ void NVDRV::Ioctl1(HLERequestContext& ctx) {
    }
    // Check device
-    std::vector<u8> output_buffer(ctx.GetWriteBufferSize(0));
+    tmp_output.resize_destructive(ctx.GetWriteBufferSize(0));
    const auto input_buffer = ctx.ReadBuffer(0);
-    const auto nv_result = nvdrv->Ioctl1(fd, command, input_buffer, output_buffer);
+    const auto nv_result = nvdrv->Ioctl1(fd, command, input_buffer, tmp_output);
    if (command.is_out != 0) {
-        ctx.WriteBuffer(output_buffer);
+        ctx.WriteBuffer(tmp_output);
    }
    IPC::ResponseBuilder rb{ctx, 3};
@@ -90,12 +90,12 @@ void NVDRV::Ioctl2(HLERequestContext& ctx) {
    const auto input_buffer = ctx.ReadBuffer(0);
    const auto input_inlined_buffer = ctx.ReadBuffer(1);
-    std::vector<u8> output_buffer(ctx.GetWriteBufferSize(0));
+    tmp_output.resize_destructive(ctx.GetWriteBufferSize(0));
    const auto nv_result =
-        nvdrv->Ioctl2(fd, command, input_buffer, input_inlined_buffer, output_buffer);
+        nvdrv->Ioctl2(fd, command, input_buffer, input_inlined_buffer, tmp_output);
    if (command.is_out != 0) {
-        ctx.WriteBuffer(output_buffer);
+        ctx.WriteBuffer(tmp_output);
    }
    IPC::ResponseBuilder rb{ctx, 3};
@@ -116,14 +116,12 @@ void NVDRV::Ioctl3(HLERequestContext& ctx) {
    }
    const auto input_buffer = ctx.ReadBuffer(0);
-    std::vector<u8> output_buffer(ctx.GetWriteBufferSize(0));
+    tmp_output.resize_destructive(ctx.GetWriteBufferSize(0));
-    std::vector<u8> output_buffer_inline(ctx.GetWriteBufferSize(1));
+    tmp_output_inline.resize_destructive(ctx.GetWriteBufferSize(1));
+    const auto nv_result = nvdrv->Ioctl3(fd, command, input_buffer, tmp_output, tmp_output_inline);
-    const auto nv_result =
-        nvdrv->Ioctl3(fd, command, input_buffer, output_buffer, output_buffer_inline);
    if (command.is_out != 0) {
-        ctx.WriteBuffer(output_buffer, 0);
+        ctx.WriteBuffer(tmp_output, 0);
-        ctx.WriteBuffer(output_buffer_inline, 1);
+        ctx.WriteBuffer(tmp_output_inline, 1);
    }
    IPC::ResponseBuilder rb{ctx, 3};
diff --git a/src/core/hle/service/nvdrv/nvdrv_interface.h b/src/core/hle/service/nvdrv/nvdrv_interface.h
index 881ea1a6b..4b593ff90 100644
--- a/src/core/hle/service/nvdrv/nvdrv_interface.h
+++ b/src/core/hle/service/nvdrv/nvdrv_interface.h
@@ -4,6 +4,7 @@
 #pragma once
 #include <memory>
+#include "common/scratch_buffer.h"
 #include "core/hle/service/nvdrv/nvdrv.h"
 #include "core/hle/service/service.h"
@@ -33,6 +34,8 @@ private:
    u64 pid{};
    bool is_initialized{};
+    Common::ScratchBuffer<u8> tmp_output;
+    Common::ScratchBuffer<u8> tmp_output_inline;
 };
 } // namespace Service::Nvidia
diff --git a/src/core/hle/service/nvnflinger/nvnflinger.cpp b/src/core/hle/service/nvnflinger/nvnflinger.cpp
index da2d5890f..5f55cd31e 100644
--- a/src/core/hle/service/nvnflinger/nvnflinger.cpp
+++ b/src/core/hle/service/nvnflinger/nvnflinger.cpp
@@ -43,14 +43,10 @@ void Nvnflinger::SplitVSync(std::stop_token stop_token) {
    Common::SetCurrentThreadPriority(Common::ThreadPriority::High);
    while (!stop_token.stop_requested()) {
-        vsync_signal.wait(false);
+        vsync_signal.Wait();
-        vsync_signal.store(false);
-        guard->lock();
+        const auto lock_guard = Lock();
        Compose();
-        guard->unlock();
    }
 }
@@ -69,8 +65,8 @@ Nvnflinger::Nvnflinger(Core::System& system_, HosBinderDriverServer& hos_binder_
        "ScreenComposition",
        [this](std::uintptr_t, s64 time,
               std::chrono::nanoseconds ns_late) -> std::optional<std::chrono::nanoseconds> {
-            vsync_signal.store(true);
+            { const auto lock_guard = Lock(); }
-            vsync_signal.notify_all();
+            vsync_signal.Set();
            return std::chrono::nanoseconds(GetNextTicks());
        });
@@ -96,8 +92,7 @@ Nvnflinger::~Nvnflinger() {
    if (system.IsMulticore()) {
        system.CoreTiming().UnscheduleEvent(multi_composition_event, {});
        vsync_thread.request_stop();
-        vsync_signal.store(true);
+        vsync_signal.Set();
-        vsync_signal.notify_all();
    } else {
        system.CoreTiming().UnscheduleEvent(single_composition_event, {});
    }
diff --git a/src/core/hle/service/nvnflinger/nvnflinger.h b/src/core/hle/service/nvnflinger/nvnflinger.h
index a043cceb2..ef236303a 100644
--- a/src/core/hle/service/nvnflinger/nvnflinger.h
+++ b/src/core/hle/service/nvnflinger/nvnflinger.h
@@ -12,6 +12,7 @@
 #include "common/common_types.h"
 #include "common/polyfill_thread.h"
+#include "common/thread.h"
 #include "core/hle/result.h"
 #include "core/hle/service/kernel_helpers.h"
@@ -143,7 +144,7 @@ private:
    Core::System& system;
-    std::atomic<bool> vsync_signal;
+    Common::Event vsync_signal;
    std::jthread vsync_thread;
diff --git a/src/core/hle/service/nvnflinger/parcel.h b/src/core/hle/service/nvnflinger/parcel.h
index fb56d75d7..23ba315a0 100644
--- a/src/core/hle/service/nvnflinger/parcel.h
+++ b/src/core/hle/service/nvnflinger/parcel.h
@@ -6,6 +6,7 @@
 #include <memory>
 #include <span>
 #include <vector>
+#include <boost/container/small_vector.hpp>
 #include "common/alignment.h"
 #include "common/assert.h"
@@ -167,7 +168,7 @@ public:
 private:
    template <typename T>
        requires(std::is_trivially_copyable_v<T>)
-    void WriteImpl(const T& val, std::vector<u8>& buffer) {
+    void WriteImpl(const T& val, boost::container::small_vector<u8, 0x200>& buffer) {
        const size_t aligned_size = Common::AlignUp(sizeof(T), 4);
        const size_t old_size = buffer.size();
        buffer.resize(old_size + aligned_size);
@@ -176,8 +177,8 @@ private:
    }
 private:
-    std::vector<u8> m_data_buffer;
+    boost::container::small_vector<u8, 0x200> m_data_buffer;
-    std::vector<u8> m_object_buffer;
+    boost::container::small_vector<u8, 0x200> m_object_buffer;
 };
 } // namespace Service::android
diff --git a/src/core/hle/service/server_manager.cpp b/src/core/hle/service/server_manager.cpp
index 156bc27d8..d1e99b184 100644
--- a/src/core/hle/service/server_manager.cpp
+++ b/src/core/hle/service/server_manager.cpp
@@ -44,7 +44,7 @@ ServerManager::~ServerManager() {
    m_event->Signal();
    // Wait for processing to stop.
-    m_stopped.wait(false);
+    m_stopped.Wait();
    m_threads.clear();
    // Clean up ports.
@@ -182,10 +182,7 @@ void ServerManager::StartAdditionalHostThreads(const char* name, size_t num_thre
 }
 Result ServerManager::LoopProcess() {
-    SCOPE_EXIT({
+    SCOPE_EXIT({ m_stopped.Set(); });
-        m_stopped.store(true);
-        m_stopped.notify_all();
-    });
    R_RETURN(this->LoopProcessImpl());
 }
diff --git a/src/core/hle/service/server_manager.h b/src/core/hle/service/server_manager.h
index fdb8af2ff..58b0a0832 100644
--- a/src/core/hle/service/server_manager.h
+++ b/src/core/hle/service/server_manager.h
@@ -3,7 +3,6 @@
 #pragma once
-#include <atomic>
 #include <functional>
 #include <list>
 #include <map>
@@ -12,6 +11,7 @@
 #include <vector>
 #include "common/polyfill_thread.h"
+#include "common/thread.h"
 #include "core/hle/result.h"
 #include "core/hle/service/mutex.h"
@@ -82,7 +82,7 @@ private:
    std::list<RequestState> m_deferrals{};
    // Host state tracking
-    std::atomic<bool> m_stopped{};
+    Common::Event m_stopped{};
    std::vector<std::jthread> m_threads{};
    std::stop_source m_stop_source{};
 };
diff --git a/src/core/hle/service/time/clock_types.h b/src/core/hle/service/time/clock_types.h
index e6293ffb9..9fc01ea90 100644
--- a/src/core/hle/service/time/clock_types.h
+++ b/src/core/hle/service/time/clock_types.h
@@ -3,6 +3,8 @@
 #pragma once
+#include <ratio>
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/uuid.h"
@@ -74,18 +76,19 @@ static_assert(std::is_trivially_copyable_v<ContinuousAdjustmentTimePoint>,
 /// https://switchbrew.org/wiki/Glue_services#TimeSpanType
 struct TimeSpanType {
    s64 nanoseconds{};
-    static constexpr s64 ns_per_second{1000000000ULL};
    s64 ToSeconds() const {
-        return nanoseconds / ns_per_second;
+        return nanoseconds / std::nano::den;
    }
    static TimeSpanType FromSeconds(s64 seconds) {
-        return {seconds * ns_per_second};
+        return {seconds * std::nano::den};
    }
-    static TimeSpanType FromTicks(u64 ticks, u64 frequency) {
+    template <u64 Frequency>
-        return FromSeconds(static_cast<s64>(ticks) / static_cast<s64>(frequency));
+    static TimeSpanType FromTicks(u64 ticks) {
+        using TicksToNSRatio = std::ratio<std::nano::den, Frequency>;
+        return {static_cast<s64>(ticks * TicksToNSRatio::num / TicksToNSRatio::den)};
    }
 };
 static_assert(sizeof(TimeSpanType) == 8, "TimeSpanType is incorrect size");
diff --git a/src/core/hle/service/time/standard_steady_clock_core.cpp b/src/core/hle/service/time/standard_steady_clock_core.cpp
index 3dbbb9850..5627b7003 100644
--- a/src/core/hle/service/time/standard_steady_clock_core.cpp
+++ b/src/core/hle/service/time/standard_steady_clock_core.cpp
@@ -10,7 +10,7 @@ namespace Service::Time::Clock {
 TimeSpanType StandardSteadyClockCore::GetCurrentRawTimePoint(Core::System& system) {
    const TimeSpanType ticks_time_span{
-        TimeSpanType::FromTicks(system.CoreTiming().GetClockTicks(), Core::Hardware::CNTFREQ)};
+        TimeSpanType::FromTicks<Core::Hardware::CNTFREQ>(system.CoreTiming().GetClockTicks())};
    TimeSpanType raw_time_point{setup_value.nanoseconds + ticks_time_span.nanoseconds};
    if (raw_time_point.nanoseconds < cached_raw_time_point.nanoseconds) {
diff --git a/src/core/hle/service/time/tick_based_steady_clock_core.cpp b/src/core/hle/service/time/tick_based_steady_clock_core.cpp
index 27600413e..0d9fb3143 100644
--- a/src/core/hle/service/time/tick_based_steady_clock_core.cpp
+++ b/src/core/hle/service/time/tick_based_steady_clock_core.cpp
@@ -10,7 +10,7 @@ namespace Service::Time::Clock {
 SteadyClockTimePoint TickBasedSteadyClockCore::GetTimePoint(Core::System& system) {
    const TimeSpanType ticks_time_span{
-        TimeSpanType::FromTicks(system.CoreTiming().GetClockTicks(), Core::Hardware::CNTFREQ)};
+        TimeSpanType::FromTicks<Core::Hardware::CNTFREQ>(system.CoreTiming().GetClockTicks())};
    return {ticks_time_span.ToSeconds(), GetClockSourceId()};
 }
diff --git a/src/core/hle/service/time/time.cpp b/src/core/hle/service/time/time.cpp
index 868be60c5..7197ca30f 100644
--- a/src/core/hle/service/time/time.cpp
+++ b/src/core/hle/service/time/time.cpp
@@ -240,8 +240,8 @@ void Module::Interface::CalculateMonotonicSystemClockBaseTimePoint(HLERequestCon
    const auto current_time_point{steady_clock_core.GetCurrentTimePoint(system)};
    if (current_time_point.clock_source_id == context.steady_time_point.clock_source_id) {
-        const auto ticks{Clock::TimeSpanType::FromTicks(system.CoreTiming().GetClockTicks(),
+        const auto ticks{Clock::TimeSpanType::FromTicks<Core::Hardware::CNTFREQ>(
-                                                        Core::Hardware::CNTFREQ)};
+            system.CoreTiming().GetClockTicks())};
        const s64 base_time_point{context.offset + current_time_point.time_point -
                                  ticks.ToSeconds()};
        IPC::ResponseBuilder rb{ctx, (sizeof(s64) / 4) + 2};
diff --git a/src/core/hle/service/time/time_sharedmemory.cpp b/src/core/hle/service/time/time_sharedmemory.cpp
index ce1c85bcc..a00676669 100644
--- a/src/core/hle/service/time/time_sharedmemory.cpp
+++ b/src/core/hle/service/time/time_sharedmemory.cpp
@@ -21,8 +21,9 @@ SharedMemory::~SharedMemory() = default;
 void SharedMemory::SetupStandardSteadyClock(const Common::UUID& clock_source_id,
                                            Clock::TimeSpanType current_time_point) {
-    const Clock::TimeSpanType ticks_time_span{Clock::TimeSpanType::FromTicks(
+    const Clock::TimeSpanType ticks_time_span{
-        system.CoreTiming().GetClockTicks(), Core::Hardware::CNTFREQ)};
+        Clock::TimeSpanType::FromTicks<Core::Hardware::CNTFREQ>(
+            system.CoreTiming().GetClockTicks())};
    const Clock::SteadyClockContext context{
        static_cast<u64>(current_time_point.nanoseconds - ticks_time_span.nanoseconds),
        clock_source_id};
diff --git a/src/core/hle/service/time/time_zone_manager.cpp b/src/core/hle/service/time/time_zone_manager.cpp
index 63aacd19f..205371a26 100644
--- a/src/core/hle/service/time/time_zone_manager.cpp
+++ b/src/core/hle/service/time/time_zone_manager.cpp
@@ -911,9 +911,13 @@ static Result ToCalendarTimeInternal(const TimeZoneRule& rules, s64 time,
    calendar_additional_info.is_dst = rules.ttis[tti_index].is_dst;
    const char* time_zone{&rules.chars[rules.ttis[tti_index].abbreviation_list_index]};
-    for (int index{}; time_zone[index] != '\0'; ++index) {
+    u32 index;
+    for (index = 0; time_zone[index] != '\0' && time_zone[index] != ',' &&
+                    index < calendar_additional_info.timezone_name.size() - 1;
+         ++index) {
        calendar_additional_info.timezone_name[index] = time_zone[index];
    }
+    calendar_additional_info.timezone_name[index] = '\0';
    return ResultSuccess;
 }
diff --git a/src/input_common/drivers/joycon.cpp b/src/input_common/drivers/joycon.cpp
index b2b5677c8..52494e0d9 100644
--- a/src/input_common/drivers/joycon.cpp
+++ b/src/input_common/drivers/joycon.cpp
@@ -195,8 +195,8 @@ void Joycons::RegisterNewDevice(SDL_hid_device_info* device_info) {
                OnMotionUpdate(port, type, id, value);
            }},
            .on_ring_data = {[this](f32 ring_data) { OnRingConUpdate(ring_data); }},
-            .on_amiibo_data = {[this, port, type](const std::vector<u8>& amiibo_data) {
+            .on_amiibo_data = {[this, port, type](const Joycon::TagInfo& tag_info) {
-                OnAmiiboUpdate(port, type, amiibo_data);
+                OnAmiiboUpdate(port, type, tag_info);
            }},
            .on_camera_data = {[this, port](const std::vector<u8>& camera_data,
                                            Joycon::IrsResolution format) {
@@ -291,13 +291,105 @@ Common::Input::NfcState Joycons::SupportsNfc(const PadIdentifier& identifier_) c
    return Common::Input::NfcState::Success;
 };
+Common::Input::NfcState Joycons::StartNfcPolling(const PadIdentifier& identifier) {
+    auto handle = GetHandle(identifier);
+    if (handle == nullptr) {
+        return Common::Input::NfcState::Unknown;
+    }
+    return TranslateDriverResult(handle->StartNfcPolling());
+};
+Common::Input::NfcState Joycons::StopNfcPolling(const PadIdentifier& identifier) {
+    auto handle = GetHandle(identifier);
+    if (handle == nullptr) {
+        return Common::Input::NfcState::Unknown;
+    }
+    return TranslateDriverResult(handle->StopNfcPolling());
+};
+Common::Input::NfcState Joycons::ReadAmiiboData(const PadIdentifier& identifier,
+                                                std::vector<u8>& out_data) {
+    auto handle = GetHandle(identifier);
+    if (handle == nullptr) {
+        return Common::Input::NfcState::Unknown;
+    }
+    return TranslateDriverResult(handle->ReadAmiiboData(out_data));
+}
 Common::Input::NfcState Joycons::WriteNfcData(const PadIdentifier& identifier,
                                              const std::vector<u8>& data) {
    auto handle = GetHandle(identifier);
-    if (handle->WriteNfcData(data) != Joycon::DriverResult::Success) {
+    if (handle == nullptr) {
-        return Common::Input::NfcState::WriteFailed;
+        return Common::Input::NfcState::Unknown;
    }
-    return Common::Input::NfcState::Success;
+    return TranslateDriverResult(handle->WriteNfcData(data));
+};
+Common::Input::NfcState Joycons::ReadMifareData(const PadIdentifier& identifier,
+                                                const Common::Input::MifareRequest& request,
+                                                Common::Input::MifareRequest& data) {
+    auto handle = GetHandle(identifier);
+    if (handle == nullptr) {
+        return Common::Input::NfcState::Unknown;
+    }
+    const auto command = static_cast<Joycon::MifareCmd>(request.data[0].command);
+    std::vector<Joycon::MifareReadChunk> read_request{};
+    for (const auto& request_data : request.data) {
+        if (request_data.command == 0) {
+            continue;
+        }
+        Joycon::MifareReadChunk chunk = {
+            .command = command,
+            .sector_key = {},
+            .sector = request_data.sector,
+        };
+        memcpy(chunk.sector_key.data(), request_data.key.data(),
+               sizeof(Joycon::MifareReadChunk::sector_key));
+        read_request.emplace_back(chunk);
+    }
+    std::vector<Joycon::MifareReadData> read_data(read_request.size());
+    const auto result = handle->ReadMifareData(read_request, read_data);
+    if (result == Joycon::DriverResult::Success) {
+        for (std::size_t i = 0; i < read_request.size(); i++) {
+            data.data[i] = {
+                .command = static_cast<u8>(command),
+                .sector = read_data[i].sector,
+                .key = {},
+                .data = read_data[i].data,
+            };
+        }
+    }
+    return TranslateDriverResult(result);
+};
+Common::Input::NfcState Joycons::WriteMifareData(const PadIdentifier& identifier,
+                                                 const Common::Input::MifareRequest& request) {
+    auto handle = GetHandle(identifier);
+    if (handle == nullptr) {
+        return Common::Input::NfcState::Unknown;
+    }
+    const auto command = static_cast<Joycon::MifareCmd>(request.data[0].command);
+    std::vector<Joycon::MifareWriteChunk> write_request{};
+    for (const auto& request_data : request.data) {
+        if (request_data.command == 0) {
+            continue;
+        }
+        Joycon::MifareWriteChunk chunk = {
+            .command = command,
+            .sector_key = {},
+            .sector = request_data.sector,
+            .data = {},
+        };
+        memcpy(chunk.sector_key.data(), request_data.key.data(),
+               sizeof(Joycon::MifareReadChunk::sector_key));
+        memcpy(chunk.data.data(), request_data.data.data(), sizeof(Joycon::MifareWriteChunk::data));
+        write_request.emplace_back(chunk);
+    }
+    return TranslateDriverResult(handle->WriteMifareData(write_request));
 };
 Common::Input::DriverResult Joycons::SetPollingMode(const PadIdentifier& identifier,
@@ -403,11 +495,20 @@ void Joycons::OnRingConUpdate(f32 ring_data) {
 }
 void Joycons::OnAmiiboUpdate(std::size_t port, Joycon::ControllerType type,
-                             const std::vector<u8>& amiibo_data) {
+                             const Joycon::TagInfo& tag_info) {
    const auto identifier = GetIdentifier(port, type);
-    const auto nfc_state = amiibo_data.empty() ? Common::Input::NfcState::AmiiboRemoved
+    const auto nfc_state = tag_info.uuid_length == 0 ? Common::Input::NfcState::AmiiboRemoved
-                                               : Common::Input::NfcState::NewAmiibo;
+                                                     : Common::Input::NfcState::NewAmiibo;
-    SetNfc(identifier, {nfc_state, amiibo_data});
+    const Common::Input::NfcStatus nfc_status{
+        .state = nfc_state,
+        .uuid_length = tag_info.uuid_length,
+        .protocol = tag_info.protocol,
+        .tag_type = tag_info.tag_type,
+        .uuid = tag_info.uuid,
+    };
+    SetNfc(identifier, nfc_status);
 }
 void Joycons::OnCameraUpdate(std::size_t port, const std::vector<u8>& camera_data,
@@ -726,4 +827,18 @@ std::string Joycons::JoyconName(Joycon::ControllerType type) const {
        return "Unknown Switch Controller";
    }
 }
+Common::Input::NfcState Joycons::TranslateDriverResult(Joycon::DriverResult result) const {
+    switch (result) {
+    case Joycon::DriverResult::Success:
+        return Common::Input::NfcState::Success;
+    case Joycon::DriverResult::Disabled:
+        return Common::Input::NfcState::WrongDeviceState;
+    case Joycon::DriverResult::NotSupported:
+        return Common::Input::NfcState::NotSupported;
+    default:
+        return Common::Input::NfcState::Unknown;
+    }
+}
 } // namespace InputCommon
diff --git a/src/input_common/drivers/joycon.h b/src/input_common/drivers/joycon.h
index e3f0ad78f..4c323d7d6 100644
--- a/src/input_common/drivers/joycon.h
+++ b/src/input_common/drivers/joycon.h
@@ -15,6 +15,7 @@ using SerialNumber = std::array<u8, 15>;
 struct Battery;
 struct Color;
 struct MotionData;
+struct TagInfo;
 enum class ControllerType : u8;
 enum class DriverResult;
 enum class IrsResolution;
@@ -39,9 +40,18 @@ public:
    Common::Input::DriverResult SetCameraFormat(const PadIdentifier& identifier,
                                                Common::Input::CameraFormat camera_format) override;
-    Common::Input::NfcState SupportsNfc(const PadIdentifier& identifier_) const override;
+    Common::Input::NfcState SupportsNfc(const PadIdentifier& identifier) const override;
-    Common::Input::NfcState WriteNfcData(const PadIdentifier& identifier_,
+    Common::Input::NfcState StartNfcPolling(const PadIdentifier& identifier) override;
+    Common::Input::NfcState StopNfcPolling(const PadIdentifier& identifier) override;
+    Common::Input::NfcState ReadAmiiboData(const PadIdentifier& identifier,
+                                           std::vector<u8>& out_data) override;
+    Common::Input::NfcState WriteNfcData(const PadIdentifier& identifier,
                                         const std::vector<u8>& data) override;
+    Common::Input::NfcState ReadMifareData(const PadIdentifier& identifier,
+                                           const Common::Input::MifareRequest& request,
+                                           Common::Input::MifareRequest& out_data) override;
+    Common::Input::NfcState WriteMifareData(const PadIdentifier& identifier,
+                                            const Common::Input::MifareRequest& request) override;
    Common::Input::DriverResult SetPollingMode(
        const PadIdentifier& identifier, const Common::Input::PollingMode polling_mode) override;
@@ -82,7 +92,7 @@ private:
                        const Joycon::MotionData& value);
    void OnRingConUpdate(f32 ring_data);
    void OnAmiiboUpdate(std::size_t port, Joycon::ControllerType type,
-                        const std::vector<u8>& amiibo_data);
+                        const Joycon::TagInfo& amiibo_data);
    void OnCameraUpdate(std::size_t port, const std::vector<u8>& camera_data,
                        Joycon::IrsResolution format);
@@ -102,6 +112,8 @@ private:
    /// Returns the name of the device in text format
    std::string JoyconName(Joycon::ControllerType type) const;
+    Common::Input::NfcState TranslateDriverResult(Joycon::DriverResult result) const;
    std::jthread scan_thread;
    // Joycon types are split by type to ease supporting dualjoycon configurations
diff --git a/src/input_common/drivers/virtual_amiibo.cpp b/src/input_common/drivers/virtual_amiibo.cpp
index 6435b8af8..180eb53ef 100644
--- a/src/input_common/drivers/virtual_amiibo.cpp
+++ b/src/input_common/drivers/virtual_amiibo.cpp
@@ -29,14 +29,13 @@ Common::Input::DriverResult VirtualAmiibo::SetPollingMode(
    switch (polling_mode) {
    case Common::Input::PollingMode::NFC:
-        if (state == State::Initialized) {
+        state = State::Initialized;
-            state = State::WaitingForAmiibo;
-        }
        return Common::Input::DriverResult::Success;
    default:
-        if (state == State::AmiiboIsOpen) {
+        if (state == State::TagNearby) {
            CloseAmiibo();
        }
+        state = State::Disabled;
        return Common::Input::DriverResult::NotSupported;
    }
 }
@@ -45,6 +44,39 @@ Common::Input::NfcState VirtualAmiibo::SupportsNfc(
    [[maybe_unused]] const PadIdentifier& identifier_) const {
    return Common::Input::NfcState::Success;
 }
+Common::Input::NfcState VirtualAmiibo::StartNfcPolling(const PadIdentifier& identifier_) {
+    if (state != State::Initialized) {
+        return Common::Input::NfcState::WrongDeviceState;
+    }
+    state = State::WaitingForAmiibo;
+    return Common::Input::NfcState::Success;
+}
+Common::Input::NfcState VirtualAmiibo::StopNfcPolling(const PadIdentifier& identifier_) {
+    if (state == State::Disabled) {
+        return Common::Input::NfcState::WrongDeviceState;
+    }
+    if (state == State::TagNearby) {
+        CloseAmiibo();
+    }
+    state = State::Initialized;
+    return Common::Input::NfcState::Success;
+}
+Common::Input::NfcState VirtualAmiibo::ReadAmiiboData(const PadIdentifier& identifier_,
+                                                      std::vector<u8>& out_data) {
+    if (state != State::TagNearby) {
+        return Common::Input::NfcState::WrongDeviceState;
+    }
+    if (status.tag_type != 1U << 1) {
+        return Common::Input::NfcState::InvalidTagType;
+    }
+    out_data.resize(nfc_data.size());
+    memcpy(out_data.data(), nfc_data.data(), nfc_data.size());
+    return Common::Input::NfcState::Success;
+}
 Common::Input::NfcState VirtualAmiibo::WriteNfcData(
    [[maybe_unused]] const PadIdentifier& identifier_, const std::vector<u8>& data) {
@@ -66,6 +98,69 @@ Common::Input::NfcState VirtualAmiibo::WriteNfcData(
    return Common::Input::NfcState::Success;
 }
+Common::Input::NfcState VirtualAmiibo::ReadMifareData(const PadIdentifier& identifier_,
+                                                      const Common::Input::MifareRequest& request,
+                                                      Common::Input::MifareRequest& out_data) {
+    if (state != State::TagNearby) {
+        return Common::Input::NfcState::WrongDeviceState;
+    }
+    if (status.tag_type != 1U << 6) {
+        return Common::Input::NfcState::InvalidTagType;
+    }
+    for (std::size_t i = 0; i < request.data.size(); i++) {
+        if (request.data[i].command == 0) {
+            continue;
+        }
+        out_data.data[i].command = request.data[i].command;
+        out_data.data[i].sector = request.data[i].sector;
+        const std::size_t sector_index =
+            request.data[i].sector * sizeof(Common::Input::MifareData::data);
+        if (nfc_data.size() < sector_index + sizeof(Common::Input::MifareData::data)) {
+            return Common::Input::NfcState::WriteFailed;
+        }
+        // Ignore the sector key as we don't support it
+        memcpy(out_data.data[i].data.data(), nfc_data.data() + sector_index,
+               sizeof(Common::Input::MifareData::data));
+    }
+    return Common::Input::NfcState::Success;
+}
+Common::Input::NfcState VirtualAmiibo::WriteMifareData(
+    const PadIdentifier& identifier_, const Common::Input::MifareRequest& request) {
+    if (state != State::TagNearby) {
+        return Common::Input::NfcState::WrongDeviceState;
+    }
+    if (status.tag_type != 1U << 6) {
+        return Common::Input::NfcState::InvalidTagType;
+    }
+    for (std::size_t i = 0; i < request.data.size(); i++) {
+        if (request.data[i].command == 0) {
+            continue;
+        }
+        const std::size_t sector_index =
+            request.data[i].sector * sizeof(Common::Input::MifareData::data);
+        if (nfc_data.size() < sector_index + sizeof(Common::Input::MifareData::data)) {
+            return Common::Input::NfcState::WriteFailed;
+        }
+        // Ignore the sector key as we don't support it
+        memcpy(nfc_data.data() + sector_index, request.data[i].data.data(),
+               sizeof(Common::Input::MifareData::data));
+    }
+    return Common::Input::NfcState::Success;
+}
 VirtualAmiibo::State VirtualAmiibo::GetCurrentState() const {
    return state;
 }
@@ -112,23 +207,31 @@ VirtualAmiibo::Info VirtualAmiibo::LoadAmiibo(std::span<u8> data) {
    case AmiiboSizeWithoutPassword:
    case AmiiboSizeWithSignature:
        nfc_data.resize(AmiiboSize);
+        status.tag_type = 1U << 1;
+        status.uuid_length = 7;
        break;
    case MifareSize:
        nfc_data.resize(MifareSize);
+        status.tag_type = 1U << 6;
+        status.uuid_length = 4;
        break;
    default:
        return Info::NotAnAmiibo;
    }
-    state = State::AmiiboIsOpen;
+    status.uuid = {};
+    status.protocol = 1;
+    state = State::TagNearby;
+    status.state = Common::Input::NfcState::NewAmiibo,
    memcpy(nfc_data.data(), data.data(), data.size_bytes());
-    SetNfc(identifier, {Common::Input::NfcState::NewAmiibo, nfc_data});
+    memcpy(status.uuid.data(), nfc_data.data(), status.uuid_length);
+    SetNfc(identifier, status);
    return Info::Success;
 }
 VirtualAmiibo::Info VirtualAmiibo::ReloadAmiibo() {
-    if (state == State::AmiiboIsOpen) {
+    if (state == State::TagNearby) {
-        SetNfc(identifier, {Common::Input::NfcState::NewAmiibo, nfc_data});
+        SetNfc(identifier, status);
        return Info::Success;
    }
@@ -136,9 +239,14 @@ VirtualAmiibo::Info VirtualAmiibo::ReloadAmiibo() {
 }
 VirtualAmiibo::Info VirtualAmiibo::CloseAmiibo() {
-    state = polling_mode == Common::Input::PollingMode::NFC ? State::WaitingForAmiibo
+    if (state != State::TagNearby) {
-                                                            : State::Initialized;
+        return Info::Success;
-    SetNfc(identifier, {Common::Input::NfcState::AmiiboRemoved, {}});
+    }
+    state = State::WaitingForAmiibo;
+    status.state = Common::Input::NfcState::AmiiboRemoved;
+    SetNfc(identifier, status);
+    status.tag_type = 0;
    return Info::Success;
 }
diff --git a/src/input_common/drivers/virtual_amiibo.h b/src/input_common/drivers/virtual_amiibo.h
index 09ca09e68..490f38e05 100644
--- a/src/input_common/drivers/virtual_amiibo.h
+++ b/src/input_common/drivers/virtual_amiibo.h
@@ -20,9 +20,10 @@ namespace InputCommon {
 class VirtualAmiibo final : public InputEngine {
 public:
    enum class State {
+        Disabled,
        Initialized,
        WaitingForAmiibo,
-        AmiiboIsOpen,
+        TagNearby,
    };
    enum class Info {
@@ -41,9 +42,17 @@ public:
        const PadIdentifier& identifier_, const Common::Input::PollingMode polling_mode_) override;
    Common::Input::NfcState SupportsNfc(const PadIdentifier& identifier_) const override;
+    Common::Input::NfcState StartNfcPolling(const PadIdentifier& identifier_) override;
+    Common::Input::NfcState StopNfcPolling(const PadIdentifier& identifier_) override;
+    Common::Input::NfcState ReadAmiiboData(const PadIdentifier& identifier_,
+                                           std::vector<u8>& out_data) override;
    Common::Input::NfcState WriteNfcData(const PadIdentifier& identifier_,
                                         const std::vector<u8>& data) override;
+    Common::Input::NfcState ReadMifareData(const PadIdentifier& identifier_,
+                                           const Common::Input::MifareRequest& data,
+                                           Common::Input::MifareRequest& out_data) override;
+    Common::Input::NfcState WriteMifareData(const PadIdentifier& identifier_,
+                                            const Common::Input::MifareRequest& data) override;
    State GetCurrentState() const;
@@ -61,8 +70,9 @@ private:
    static constexpr std::size_t MifareSize = 0x400;
    std::string file_path{};
-    State state{State::Initialized};
+    State state{State::Disabled};
    std::vector<u8> nfc_data;
+    Common::Input::NfcStatus status;
    Common::Input::PollingMode polling_mode{Common::Input::PollingMode::Passive};
 };
 } // namespace InputCommon
diff --git a/src/input_common/helpers/joycon_driver.cpp b/src/input_common/helpers/joycon_driver.cpp
index 95106f16d..2c8c66951 100644
--- a/src/input_common/helpers/joycon_driver.cpp
+++ b/src/input_common/helpers/joycon_driver.cpp
@@ -2,6 +2,7 @@
 // SPDX-License-Identifier: GPL-2.0-or-later
 #include "common/logging/log.h"
+#include "common/scope_exit.h"
 #include "common/swap.h"
 #include "common/thread.h"
 #include "input_common/helpers/joycon_driver.h"
@@ -112,7 +113,7 @@ DriverResult JoyconDriver::InitializeDevice() {
    joycon_poller = std::make_unique<JoyconPoller>(device_type, left_stick_calibration,
                                                   right_stick_calibration, motion_calibration);
-    // Start pooling for data
+    // Start polling for data
    is_connected = true;
    if (!input_thread_running) {
        input_thread =
@@ -208,7 +209,7 @@ void JoyconDriver::OnNewData(std::span<u8> buffer) {
        joycon_poller->UpdateCamera(irs_protocol->GetImage(), irs_protocol->GetIrsFormat());
    }
-    if (nfc_protocol->IsEnabled()) {
+    if (nfc_protocol->IsPolling()) {
        if (amiibo_detected) {
            if (!nfc_protocol->HasAmiibo()) {
                joycon_poller->UpdateAmiibo({});
@@ -218,10 +219,10 @@ void JoyconDriver::OnNewData(std::span<u8> buffer) {
        }
        if (!amiibo_detected) {
-            std::vector<u8> data(0x21C);
+            Joycon::TagInfo tag_info;
-            const auto result = nfc_protocol->ScanAmiibo(data);
+            const auto result = nfc_protocol->GetTagInfo(tag_info);
            if (result == DriverResult::Success) {
-                joycon_poller->UpdateAmiibo(data);
+                joycon_poller->UpdateAmiibo(tag_info);
                amiibo_detected = true;
            }
        }
@@ -247,6 +248,7 @@ void JoyconDriver::OnNewData(std::span<u8> buffer) {
 }
 DriverResult JoyconDriver::SetPollingMode() {
+    SCOPE_EXIT({ disable_input_thread = false; });
    disable_input_thread = true;
    rumble_protocol->EnableRumble(vibration_enabled && supported_features.vibration);
@@ -276,7 +278,6 @@ DriverResult JoyconDriver::SetPollingMode() {
    if (irs_enabled && supported_features.irs) {
        auto result = irs_protocol->EnableIrs();
        if (result == DriverResult::Success) {
-            disable_input_thread = false;
            return result;
        }
        irs_protocol->DisableIrs();
@@ -286,10 +287,6 @@ DriverResult JoyconDriver::SetPollingMode() {
    if (nfc_enabled && supported_features.nfc) {
        auto result = nfc_protocol->EnableNfc();
        if (result == DriverResult::Success) {
-            result = nfc_protocol->StartNFCPollingMode();
-        }
-        if (result == DriverResult::Success) {
-            disable_input_thread = false;
            return result;
        }
        nfc_protocol->DisableNfc();
@@ -303,7 +300,6 @@ DriverResult JoyconDriver::SetPollingMode() {
        }
        if (result == DriverResult::Success) {
            ring_connected = true;
-            disable_input_thread = false;
            return result;
        }
        ring_connected = false;
@@ -314,7 +310,6 @@ DriverResult JoyconDriver::SetPollingMode() {
    if (passive_enabled && supported_features.passive) {
        const auto result = generic_protocol->EnablePassiveMode();
        if (result == DriverResult::Success) {
-            disable_input_thread = false;
            return result;
        }
        LOG_ERROR(Input, "Error enabling passive mode");
@@ -328,7 +323,6 @@ DriverResult JoyconDriver::SetPollingMode() {
    // Switch calls this function after enabling active mode
    generic_protocol->TriggersElapsed();
-    disable_input_thread = false;
    return result;
 }
@@ -492,9 +486,63 @@ DriverResult JoyconDriver::SetRingConMode() {
    return result;
 }
-DriverResult JoyconDriver::WriteNfcData(std::span<const u8> data) {
+DriverResult JoyconDriver::StartNfcPolling() {
    std::scoped_lock lock{mutex};
+    if (!supported_features.nfc) {
+        return DriverResult::NotSupported;
+    }
+    if (!nfc_protocol->IsEnabled()) {
+        return DriverResult::Disabled;
+    }
+    disable_input_thread = true;
+    const auto result = nfc_protocol->StartNFCPollingMode();
+    disable_input_thread = false;
+    return result;
+}
+DriverResult JoyconDriver::StopNfcPolling() {
+    std::scoped_lock lock{mutex};
+    if (!supported_features.nfc) {
+        return DriverResult::NotSupported;
+    }
+    if (!nfc_protocol->IsEnabled()) {
+        return DriverResult::Disabled;
+    }
+    disable_input_thread = true;
+    const auto result = nfc_protocol->StopNFCPollingMode();
+    disable_input_thread = false;
+    return result;
+}
+DriverResult JoyconDriver::ReadAmiiboData(std::vector<u8>& out_data) {
+    std::scoped_lock lock{mutex};
+    if (!supported_features.nfc) {
+        return DriverResult::NotSupported;
+    }
+    if (!nfc_protocol->IsEnabled()) {
+        return DriverResult::Disabled;
+    }
+    if (!amiibo_detected) {
+        return DriverResult::ErrorWritingData;
+    }
+    out_data.resize(0x21C);
    disable_input_thread = true;
+    const auto result = nfc_protocol->ReadAmiibo(out_data);
+    disable_input_thread = false;
+    return result;
+}
+DriverResult JoyconDriver::WriteNfcData(std::span<const u8> data) {
+    std::scoped_lock lock{mutex};
    if (!supported_features.nfc) {
        return DriverResult::NotSupported;
@@ -506,9 +554,51 @@ DriverResult JoyconDriver::WriteNfcData(std::span<const u8> data) {
        return DriverResult::ErrorWritingData;
    }
+    disable_input_thread = true;
    const auto result = nfc_protocol->WriteAmiibo(data);
+    disable_input_thread = false;
+    return result;
+}
+DriverResult JoyconDriver::ReadMifareData(std::span<const MifareReadChunk> data,
+                                          std::span<MifareReadData> out_data) {
+    std::scoped_lock lock{mutex};
+    if (!supported_features.nfc) {
+        return DriverResult::NotSupported;
+    }
+    if (!nfc_protocol->IsEnabled()) {
+        return DriverResult::Disabled;
+    }
+    if (!amiibo_detected) {
+        return DriverResult::ErrorWritingData;
+    }
+    disable_input_thread = true;
+    const auto result = nfc_protocol->ReadMifare(data, out_data);
+    disable_input_thread = false;
+    return result;
+}
+DriverResult JoyconDriver::WriteMifareData(std::span<const MifareWriteChunk> data) {
+    std::scoped_lock lock{mutex};
+    if (!supported_features.nfc) {
+        return DriverResult::NotSupported;
+    }
+    if (!nfc_protocol->IsEnabled()) {
+        return DriverResult::Disabled;
+    }
+    if (!amiibo_detected) {
+        return DriverResult::ErrorWritingData;
+    }
+    disable_input_thread = true;
+    const auto result = nfc_protocol->WriteMifare(data);
    disable_input_thread = false;
    return result;
 }
diff --git a/src/input_common/helpers/joycon_driver.h b/src/input_common/helpers/joycon_driver.h
index e9b2fccbb..bc7025a21 100644
--- a/src/input_common/helpers/joycon_driver.h
+++ b/src/input_common/helpers/joycon_driver.h
@@ -49,7 +49,13 @@ public:
    DriverResult SetIrMode();
    DriverResult SetNfcMode();
    DriverResult SetRingConMode();
+    DriverResult StartNfcPolling();
+    DriverResult StopNfcPolling();
+    DriverResult ReadAmiiboData(std::vector<u8>& out_data);
    DriverResult WriteNfcData(std::span<const u8> data);
+    DriverResult ReadMifareData(std::span<const MifareReadChunk> request,
+                                std::span<MifareReadData> out_data);
+    DriverResult WriteMifareData(std::span<const MifareWriteChunk> request);
    void SetCallbacks(const JoyconCallbacks& callbacks);
diff --git a/src/input_common/helpers/joycon_protocol/joycon_types.h b/src/input_common/helpers/joycon_protocol/joycon_types.h
index 5007b0e18..e0e431156 100644
--- a/src/input_common/helpers/joycon_protocol/joycon_types.h
+++ b/src/input_common/helpers/joycon_protocol/joycon_types.h
@@ -24,6 +24,7 @@ constexpr std::array<u8, 8> DefaultVibrationBuffer{0x0, 0x1, 0x40, 0x40, 0x0, 0x
 using MacAddress = std::array<u8, 6>;
 using SerialNumber = std::array<u8, 15>;
 using TagUUID = std::array<u8, 7>;
+using MifareUUID = std::array<u8, 4>;
 enum class ControllerType : u8 {
    None = 0x00,
@@ -307,6 +308,19 @@ enum class NFCStatus : u8 {
    WriteDone = 0x05,
    TagLost = 0x07,
    WriteReady = 0x09,
+    MifareDone = 0x10,
+};
+enum class MifareCmd : u8 {
+    None = 0x00,
+    Read = 0x30,
+    AuthA = 0x60,
+    AuthB = 0x61,
+    Write = 0xA0,
+    Transfer = 0xB0,
+    Decrement = 0xC0,
+    Increment = 0xC1,
+    Store = 0xC2
 };
 enum class IrsMode : u8 {
@@ -592,6 +606,14 @@ struct NFCWriteCommandData {
 static_assert(sizeof(NFCWriteCommandData) == 0x15, "NFCWriteCommandData is an invalid size");
 #pragma pack(pop)
+struct MifareCommandData {
+    u8 unknown1;
+    u8 unknown2;
+    u8 number_of_short_bytes;
+    MifareUUID uid;
+};
+static_assert(sizeof(MifareCommandData) == 0x7, "MifareCommandData is an invalid size");
 struct NFCPollingCommandData {
    u8 enable_mifare;
    u8 unknown_1;
@@ -629,6 +651,41 @@ struct NFCWritePackage {
    std::array<NFCDataChunk, 4> data_chunks;
 };
+struct MifareReadChunk {
+    MifareCmd command;
+    std::array<u8, 0x6> sector_key;
+    u8 sector;
+};
+struct MifareWriteChunk {
+    MifareCmd command;
+    std::array<u8, 0x6> sector_key;
+    u8 sector;
+    std::array<u8, 0x10> data;
+};
+struct MifareReadData {
+    u8 sector;
+    std::array<u8, 0x10> data;
+};
+struct MifareReadPackage {
+    MifareCommandData command_data;
+    std::array<MifareReadChunk, 0x10> data_chunks;
+};
+struct MifareWritePackage {
+    MifareCommandData command_data;
+    std::array<MifareWriteChunk, 0x10> data_chunks;
+};
+struct TagInfo {
+    u8 uuid_length;
+    u8 protocol;
+    u8 tag_type;
+    std::array<u8, 10> uuid;
+};
 struct IrsConfigure {
    MCUCommand command;
    MCUSubCommand sub_command;
@@ -744,7 +801,7 @@ struct JoyconCallbacks {
    std::function<void(int, f32)> on_stick_data;
    std::function<void(int, const MotionData&)> on_motion_data;
    std::function<void(f32)> on_ring_data;
-    std::function<void(const std::vector<u8>&)> on_amiibo_data;
+    std::function<void(const Joycon::TagInfo&)> on_amiibo_data;
    std::function<void(const std::vector<u8>&, IrsResolution)> on_camera_data;
 };
diff --git a/src/input_common/helpers/joycon_protocol/nfc.cpp b/src/input_common/helpers/joycon_protocol/nfc.cpp
index f7058c4a7..261f46255 100644
--- a/src/input_common/helpers/joycon_protocol/nfc.cpp
+++ b/src/input_common/helpers/joycon_protocol/nfc.cpp
@@ -40,6 +40,16 @@ DriverResult NfcProtocol::EnableNfc() {
    if (result == DriverResult::Success) {
        result = WaitUntilNfcIs(NFCStatus::Ready);
    }
+    if (result == DriverResult::Success) {
+        MCUCommandResponse output{};
+        result = SendStopPollingRequest(output);
+    }
+    if (result == DriverResult::Success) {
+        result = WaitUntilNfcIs(NFCStatus::Ready);
+    }
+    if (result == DriverResult::Success) {
+        is_enabled = true;
+    }
    return result;
 }
@@ -54,37 +64,50 @@ DriverResult NfcProtocol::DisableNfc() {
    }
    is_enabled = false;
+    is_polling = false;
    return result;
 }
 DriverResult NfcProtocol::StartNFCPollingMode() {
-    LOG_DEBUG(Input, "Start NFC pooling Mode");
+    LOG_DEBUG(Input, "Start NFC polling Mode");
    ScopedSetBlocking sb(this);
    DriverResult result{DriverResult::Success};
    if (result == DriverResult::Success) {
        MCUCommandResponse output{};
-        result = SendStopPollingRequest(output);
+        result = SendStartPollingRequest(output);
    }
    if (result == DriverResult::Success) {
-        result = WaitUntilNfcIs(NFCStatus::Ready);
+        result = WaitUntilNfcIs(NFCStatus::Polling);
    }
    if (result == DriverResult::Success) {
+        is_polling = true;
+    }
+    return result;
+}
+DriverResult NfcProtocol::StopNFCPollingMode() {
+    LOG_DEBUG(Input, "Stop NFC polling Mode");
+    ScopedSetBlocking sb(this);
+    DriverResult result{DriverResult::Success};
+    if (result == DriverResult::Success) {
        MCUCommandResponse output{};
-        result = SendStartPollingRequest(output);
+        result = SendStopPollingRequest(output);
    }
    if (result == DriverResult::Success) {
-        result = WaitUntilNfcIs(NFCStatus::Polling);
+        result = WaitUntilNfcIs(NFCStatus::WriteReady);
    }
    if (result == DriverResult::Success) {
-        is_enabled = true;
+        is_polling = false;
    }
    return result;
 }
-DriverResult NfcProtocol::ScanAmiibo(std::vector<u8>& data) {
+DriverResult NfcProtocol::GetTagInfo(Joycon::TagInfo& tag_info) {
    if (update_counter++ < AMIIBO_UPDATE_DELAY) {
        return DriverResult::Delayed;
    }
@@ -100,11 +123,41 @@ DriverResult NfcProtocol::ScanAmiibo(std::vector<u8>& data) {
    }
    if (result == DriverResult::Success) {
+        tag_info = {
+            .uuid_length = tag_data.uuid_size,
+            .protocol = 1,
+            .tag_type = tag_data.type,
+            .uuid = {},
+        };
+        memcpy(tag_info.uuid.data(), tag_data.uuid.data(), tag_data.uuid_size);
+        // Investigate why mifare type is not correct
+        if (tag_info.tag_type == 144) {
+            tag_info.tag_type = 1U << 6;
+        }
        std::string uuid_string;
        for (auto& content : tag_data.uuid) {
            uuid_string += fmt::format(" {:02x}", content);
        }
        LOG_INFO(Input, "Tag detected, type={}, uuid={}", tag_data.type, uuid_string);
+    }
+    return result;
+}
+DriverResult NfcProtocol::ReadAmiibo(std::vector<u8>& data) {
+    LOG_DEBUG(Input, "Scan for amiibos");
+    ScopedSetBlocking sb(this);
+    DriverResult result{DriverResult::Success};
+    TagFoundData tag_data{};
+    if (result == DriverResult::Success) {
+        result = IsTagInRange(tag_data, 7);
+    }
+    if (result == DriverResult::Success) {
        result = GetAmiiboData(data);
    }
@@ -154,6 +207,69 @@ DriverResult NfcProtocol::WriteAmiibo(std::span<const u8> data) {
    return result;
 }
+DriverResult NfcProtocol::ReadMifare(std::span<const MifareReadChunk> read_request,
+                                     std::span<MifareReadData> out_data) {
+    LOG_DEBUG(Input, "Read mifare");
+    ScopedSetBlocking sb(this);
+    DriverResult result{DriverResult::Success};
+    TagFoundData tag_data{};
+    MifareUUID tag_uuid{};
+    if (result == DriverResult::Success) {
+        result = IsTagInRange(tag_data, 7);
+    }
+    if (result == DriverResult::Success) {
+        memcpy(tag_uuid.data(), tag_data.uuid.data(), sizeof(MifareUUID));
+        result = GetMifareData(tag_uuid, read_request, out_data);
+    }
+    if (result == DriverResult::Success) {
+        MCUCommandResponse output{};
+        result = SendStopPollingRequest(output);
+    }
+    if (result == DriverResult::Success) {
+        result = WaitUntilNfcIs(NFCStatus::Ready);
+    }
+    if (result == DriverResult::Success) {
+        MCUCommandResponse output{};
+        result = SendStartPollingRequest(output, true);
+    }
+    if (result == DriverResult::Success) {
+        result = WaitUntilNfcIs(NFCStatus::WriteReady);
+    }
+    return result;
+}
+DriverResult NfcProtocol::WriteMifare(std::span<const MifareWriteChunk> write_request) {
+    LOG_DEBUG(Input, "Write mifare");
+    ScopedSetBlocking sb(this);
+    DriverResult result{DriverResult::Success};
+    TagFoundData tag_data{};
+    MifareUUID tag_uuid{};
+    if (result == DriverResult::Success) {
+        result = IsTagInRange(tag_data, 7);
+    }
+    if (result == DriverResult::Success) {
+        memcpy(tag_uuid.data(), tag_data.uuid.data(), sizeof(MifareUUID));
+        result = WriteMifareData(tag_uuid, write_request);
+    }
+    if (result == DriverResult::Success) {
+        MCUCommandResponse output{};
+        result = SendStopPollingRequest(output);
+    }
+    if (result == DriverResult::Success) {
+        result = WaitUntilNfcIs(NFCStatus::Ready);
+    }
+    if (result == DriverResult::Success) {
+        MCUCommandResponse output{};
+        result = SendStartPollingRequest(output, true);
+    }
+    if (result == DriverResult::Success) {
+        result = WaitUntilNfcIs(NFCStatus::WriteReady);
+    }
+    return result;
+}
 bool NfcProtocol::HasAmiibo() {
    if (update_counter++ < AMIIBO_UPDATE_DELAY) {
        return true;
@@ -341,6 +457,158 @@ DriverResult NfcProtocol::WriteAmiiboData(const TagUUID& tag_uuid, std::span<con
    return result;
 }
+DriverResult NfcProtocol::GetMifareData(const MifareUUID& tag_uuid,
+                                        std::span<const MifareReadChunk> read_request,
+                                        std::span<MifareReadData> out_data) {
+    constexpr std::size_t timeout_limit = 60;
+    const auto nfc_data = MakeMifareReadPackage(tag_uuid, read_request);
+    const std::vector<u8> nfc_buffer_data = SerializeMifareReadPackage(nfc_data);
+    std::span<const u8> buffer(nfc_buffer_data);
+    DriverResult result = DriverResult::Success;
+    MCUCommandResponse output{};
+    u8 block_id = 1;
+    u8 package_index = 0;
+    std::size_t tries = 0;
+    std::size_t current_position = 0;
+    LOG_INFO(Input, "Reading Mifare data");
+    // Send data request. Nfc buffer size is 31, Send the data in smaller packages
+    while (current_position < buffer.size() && tries++ < timeout_limit) {
+        const std::size_t next_position =
+            std::min(current_position + sizeof(NFCRequestState::raw_data), buffer.size());
+        const std::size_t block_size = next_position - current_position;
+        const bool is_last_packet = block_size < sizeof(NFCRequestState::raw_data);
+        SendReadDataMifareRequest(output, block_id, is_last_packet,
+                                  buffer.subspan(current_position, block_size));
+        const auto nfc_status = static_cast<NFCStatus>(output.mcu_data[6]);
+        if (output.mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::TagLost) {
+            return DriverResult::ErrorReadingData;
+        }
+        // Increase position when data is confirmed by the joycon
+        if (output.mcu_report == MCUReport::NFCState &&
+            (output.mcu_data[1] << 8) + output.mcu_data[0] == 0x0500 &&
+            output.mcu_data[3] == block_id) {
+            block_id++;
+            current_position = next_position;
+        }
+    }
+    if (result != DriverResult::Success) {
+        return result;
+    }
+    // Wait for reply and save the output data
+    while (tries++ < timeout_limit) {
+        result = SendNextPackageRequest(output, package_index);
+        const auto nfc_status = static_cast<NFCStatus>(output.mcu_data[6]);
+        if (result != DriverResult::Success) {
+            return result;
+        }
+        if (output.mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::TagLost) {
+            return DriverResult::ErrorReadingData;
+        }
+        if (output.mcu_report == MCUReport::NFCState && output.mcu_data[1] == 0x10) {
+            constexpr std::size_t DATA_LENGHT = 0x10 + 1;
+            constexpr std::size_t DATA_START = 11;
+            const u8 number_of_elements = output.mcu_data[10];
+            for (std::size_t i = 0; i < number_of_elements; i++) {
+                out_data[i].sector = output.mcu_data[DATA_START + (i * DATA_LENGHT)];
+                memcpy(out_data[i].data.data(),
+                       output.mcu_data.data() + DATA_START + 1 + (i * DATA_LENGHT),
+                       sizeof(MifareReadData::data));
+            }
+            package_index++;
+            continue;
+        }
+        if (output.mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::MifareDone) {
+            LOG_INFO(Input, "Finished reading mifare");
+            break;
+        }
+    }
+    return result;
+}
+DriverResult NfcProtocol::WriteMifareData(const MifareUUID& tag_uuid,
+                                          std::span<const MifareWriteChunk> write_request) {
+    constexpr std::size_t timeout_limit = 60;
+    const auto nfc_data = MakeMifareWritePackage(tag_uuid, write_request);
+    const std::vector<u8> nfc_buffer_data = SerializeMifareWritePackage(nfc_data);
+    std::span<const u8> buffer(nfc_buffer_data);
+    DriverResult result = DriverResult::Success;
+    MCUCommandResponse output{};
+    u8 block_id = 1;
+    u8 package_index = 0;
+    std::size_t tries = 0;
+    std::size_t current_position = 0;
+    LOG_INFO(Input, "Writing Mifare data");
+    // Send data request. Nfc buffer size is 31, Send the data in smaller packages
+    while (current_position < buffer.size() && tries++ < timeout_limit) {
+        const std::size_t next_position =
+            std::min(current_position + sizeof(NFCRequestState::raw_data), buffer.size());
+        const std::size_t block_size = next_position - current_position;
+        const bool is_last_packet = block_size < sizeof(NFCRequestState::raw_data);
+        SendReadDataMifareRequest(output, block_id, is_last_packet,
+                                  buffer.subspan(current_position, block_size));
+        const auto nfc_status = static_cast<NFCStatus>(output.mcu_data[6]);
+        if (output.mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::TagLost) {
+            return DriverResult::ErrorReadingData;
+        }
+        // Increase position when data is confirmed by the joycon
+        if (output.mcu_report == MCUReport::NFCState &&
+            (output.mcu_data[1] << 8) + output.mcu_data[0] == 0x0500 &&
+            output.mcu_data[3] == block_id) {
+            block_id++;
+            current_position = next_position;
+        }
+    }
+    if (result != DriverResult::Success) {
+        return result;
+    }
+    // Wait for reply and ignore the output data
+    while (tries++ < timeout_limit) {
+        result = SendNextPackageRequest(output, package_index);
+        const auto nfc_status = static_cast<NFCStatus>(output.mcu_data[6]);
+        if (result != DriverResult::Success) {
+            return result;
+        }
+        if (output.mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::TagLost) {
+            return DriverResult::ErrorReadingData;
+        }
+        if (output.mcu_report == MCUReport::NFCState && output.mcu_data[1] == 0x10) {
+            package_index++;
+            continue;
+        }
+        if (output.mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::MifareDone) {
+            LOG_INFO(Input, "Finished writing mifare");
+            break;
+        }
+    }
+    return result;
+}
 DriverResult NfcProtocol::SendStartPollingRequest(MCUCommandResponse& output,
                                                  bool is_second_attempt) {
    NFCRequestState request{
@@ -477,6 +745,28 @@ DriverResult NfcProtocol::SendWriteDataAmiiboRequest(MCUCommandResponse& output,
                       output);
 }
+DriverResult NfcProtocol::SendReadDataMifareRequest(MCUCommandResponse& output, u8 block_id,
+                                                    bool is_last_packet, std::span<const u8> data) {
+    const auto data_size = std::min(data.size(), sizeof(NFCRequestState::raw_data));
+    NFCRequestState request{
+        .command_argument = NFCCommand::Mifare,
+        .block_id = block_id,
+        .packet_id = {},
+        .packet_flag =
+            is_last_packet ? MCUPacketFlag::LastCommandPacket : MCUPacketFlag::MorePacketsRemaining,
+        .data_length = static_cast<u8>(data_size),
+        .raw_data = {},
+        .crc = {},
+    };
+    memcpy(request.raw_data.data(), data.data(), data_size);
+    std::array<u8, sizeof(NFCRequestState)> request_data{};
+    memcpy(request_data.data(), &request, sizeof(NFCRequestState));
+    request_data[36] = CalculateMCU_CRC8(request_data.data(), 36);
+    return SendMCUData(ReportMode::NFC_IR_MODE_60HZ, MCUSubCommand::ReadDeviceMode, request_data,
+                       output);
+}
 std::vector<u8> NfcProtocol::SerializeWritePackage(const NFCWritePackage& package) const {
    const std::size_t header_size =
        sizeof(NFCWriteCommandData) + sizeof(NFCWritePackage::number_of_chunks);
@@ -498,6 +788,48 @@ std::vector<u8> NfcProtocol::SerializeWritePackage(const NFCWritePackage& packag
    return serialized_data;
 }
+std::vector<u8> NfcProtocol::SerializeMifareReadPackage(const MifareReadPackage& package) const {
+    const std::size_t header_size = sizeof(MifareCommandData);
+    std::vector<u8> serialized_data(header_size);
+    std::size_t start_index = 0;
+    memcpy(serialized_data.data(), &package, header_size);
+    start_index += header_size;
+    for (const auto& data_chunk : package.data_chunks) {
+        const std::size_t chunk_size = sizeof(MifareReadChunk);
+        if (data_chunk.command == MifareCmd::None) {
+            continue;
+        }
+        serialized_data.resize(start_index + chunk_size);
+        memcpy(serialized_data.data() + start_index, &data_chunk, chunk_size);
+        start_index += chunk_size;
+    }
+    return serialized_data;
+}
+std::vector<u8> NfcProtocol::SerializeMifareWritePackage(const MifareWritePackage& package) const {
+    const std::size_t header_size = sizeof(MifareCommandData);
+    std::vector<u8> serialized_data(header_size);
+    std::size_t start_index = 0;
+    memcpy(serialized_data.data(), &package, header_size);
+    start_index += header_size;
+    for (const auto& data_chunk : package.data_chunks) {
+        const std::size_t chunk_size = sizeof(MifareWriteChunk);
+        if (data_chunk.command == MifareCmd::None) {
+            continue;
+        }
+        serialized_data.resize(start_index + chunk_size);
+        memcpy(serialized_data.data() + start_index, &data_chunk, chunk_size);
+        start_index += chunk_size;
+    }
+    return serialized_data;
+}
 NFCWritePackage NfcProtocol::MakeAmiiboWritePackage(const TagUUID& tag_uuid,
                                                    std::span<const u8> data) const {
    return {
@@ -527,6 +859,46 @@ NFCWritePackage NfcProtocol::MakeAmiiboWritePackage(const TagUUID& tag_uuid,
    };
 }
+MifareReadPackage NfcProtocol::MakeMifareReadPackage(
+    const MifareUUID& tag_uuid, std::span<const MifareReadChunk> read_request) const {
+    MifareReadPackage package{
+        .command_data{
+            .unknown1 = 0xd0,
+            .unknown2 = 0x07,
+            .number_of_short_bytes = static_cast<u8>(
+                ((read_request.size() * sizeof(MifareReadChunk)) + sizeof(MifareUUID)) / 2),
+            .uid = tag_uuid,
+        },
+        .data_chunks = {},
+    };
+    for (std::size_t i = 0; i < read_request.size() && i < package.data_chunks.size(); ++i) {
+        package.data_chunks[i] = read_request[i];
+    }
+    return package;
+}
+MifareWritePackage NfcProtocol::MakeMifareWritePackage(
+    const MifareUUID& tag_uuid, std::span<const MifareWriteChunk> read_request) const {
+    MifareWritePackage package{
+        .command_data{
+            .unknown1 = 0xd0,
+            .unknown2 = 0x07,
+            .number_of_short_bytes = static_cast<u8>(
+                ((read_request.size() * sizeof(MifareReadChunk)) + sizeof(MifareUUID) + 2) / 2),
+            .uid = tag_uuid,
+        },
+        .data_chunks = {},
+    };
+    for (std::size_t i = 0; i < read_request.size() && i < package.data_chunks.size(); ++i) {
+        package.data_chunks[i] = read_request[i];
+    }
+    return package;
+}
 NFCDataChunk NfcProtocol::MakeAmiiboChunk(u8 page, u8 size, std::span<const u8> data) const {
    constexpr u8 NFC_PAGE_SIZE = 4;
@@ -606,4 +978,8 @@ bool NfcProtocol::IsEnabled() const {
    return is_enabled;
 }
+bool NfcProtocol::IsPolling() const {
+    return is_polling;
+}
 } // namespace InputCommon::Joycon
diff --git a/src/input_common/helpers/joycon_protocol/nfc.h b/src/input_common/helpers/joycon_protocol/nfc.h
index eb58c427d..0be95e40e 100644
--- a/src/input_common/helpers/joycon_protocol/nfc.h
+++ b/src/input_common/helpers/joycon_protocol/nfc.h
@@ -25,14 +25,25 @@ public:
    DriverResult StartNFCPollingMode();
-    DriverResult ScanAmiibo(std::vector<u8>& data);
+    DriverResult StopNFCPollingMode();
+    DriverResult GetTagInfo(Joycon::TagInfo& tag_info);
+    DriverResult ReadAmiibo(std::vector<u8>& data);
    DriverResult WriteAmiibo(std::span<const u8> data);
+    DriverResult ReadMifare(std::span<const MifareReadChunk> read_request,
+                            std::span<MifareReadData> out_data);
+    DriverResult WriteMifare(std::span<const MifareWriteChunk> write_request);
    bool HasAmiibo();
    bool IsEnabled() const;
+    bool IsPolling() const;
 private:
    // Number of times the function will be delayed until it outputs valid data
    static constexpr std::size_t AMIIBO_UPDATE_DELAY = 15;
@@ -51,6 +62,13 @@ private:
    DriverResult WriteAmiiboData(const TagUUID& tag_uuid, std::span<const u8> data);
+    DriverResult GetMifareData(const MifareUUID& tag_uuid,
+                               std::span<const MifareReadChunk> read_request,
+                               std::span<MifareReadData> out_data);
+    DriverResult WriteMifareData(const MifareUUID& tag_uuid,
+                                 std::span<const MifareWriteChunk> write_request);
    DriverResult SendStartPollingRequest(MCUCommandResponse& output,
                                         bool is_second_attempt = false);
@@ -65,17 +83,31 @@ private:
    DriverResult SendWriteDataAmiiboRequest(MCUCommandResponse& output, u8 block_id,
                                            bool is_last_packet, std::span<const u8> data);
+    DriverResult SendReadDataMifareRequest(MCUCommandResponse& output, u8 block_id,
+                                           bool is_last_packet, std::span<const u8> data);
    std::vector<u8> SerializeWritePackage(const NFCWritePackage& package) const;
+    std::vector<u8> SerializeMifareReadPackage(const MifareReadPackage& package) const;
+    std::vector<u8> SerializeMifareWritePackage(const MifareWritePackage& package) const;
    NFCWritePackage MakeAmiiboWritePackage(const TagUUID& tag_uuid, std::span<const u8> data) const;
    NFCDataChunk MakeAmiiboChunk(u8 page, u8 size, std::span<const u8> data) const;
+    MifareReadPackage MakeMifareReadPackage(const MifareUUID& tag_uuid,
+                                            std::span<const MifareReadChunk> read_request) const;
+    MifareWritePackage MakeMifareWritePackage(const MifareUUID& tag_uuid,
+                                              std::span<const MifareWriteChunk> read_request) const;
    NFCReadBlockCommand GetReadBlockCommand(NFCPages pages) const;
    TagUUID GetTagUUID(std::span<const u8> data) const;
    bool is_enabled{};
+    bool is_polling{};
    std::size_t update_counter{};
 };
diff --git a/src/input_common/helpers/joycon_protocol/poller.cpp b/src/input_common/helpers/joycon_protocol/poller.cpp
index dca797f7a..1aab9e12a 100644
--- a/src/input_common/helpers/joycon_protocol/poller.cpp
+++ b/src/input_common/helpers/joycon_protocol/poller.cpp
@@ -70,8 +70,8 @@ void JoyconPoller::UpdateColor(const Color& color) {
    callbacks.on_color_data(color);
 }
-void JoyconPoller::UpdateAmiibo(const std::vector<u8>& amiibo_data) {
+void JoyconPoller::UpdateAmiibo(const Joycon::TagInfo& tag_info) {
-    callbacks.on_amiibo_data(amiibo_data);
+    callbacks.on_amiibo_data(tag_info);
 }
 void JoyconPoller::UpdateCamera(const std::vector<u8>& camera_data, IrsResolution format) {
diff --git a/src/input_common/helpers/joycon_protocol/poller.h b/src/input_common/helpers/joycon_protocol/poller.h
index 0fa72c6db..3746abe5d 100644
--- a/src/input_common/helpers/joycon_protocol/poller.h
+++ b/src/input_common/helpers/joycon_protocol/poller.h
@@ -36,8 +36,8 @@ public:
    void UpdateColor(const Color& color);
    void UpdateRing(s16 value, const RingStatus& ring_status);
-    void UpdateAmiibo(const std::vector<u8>& amiibo_data);
+    void UpdateAmiibo(const Joycon::TagInfo& tag_info);
-    void UpdateCamera(const std::vector<u8>& amiibo_data, IrsResolution format);
+    void UpdateCamera(const std::vector<u8>& camera_data, IrsResolution format);
 private:
    void UpdateActiveLeftPadInput(const InputReportActive& input,
diff --git a/src/input_common/input_engine.h b/src/input_common/input_engine.h
index 50b5a3dc8..c2d0cbb34 100644
--- a/src/input_common/input_engine.h
+++ b/src/input_common/input_engine.h
@@ -143,12 +143,46 @@ public:
        return Common::Input::NfcState::NotSupported;
    }
+    // Start scanning for nfc tags
+    virtual Common::Input::NfcState StartNfcPolling(
+        [[maybe_unused]] const PadIdentifier& identifier_) {
+        return Common::Input::NfcState::NotSupported;
+    }
+    // Start scanning for nfc tags
+    virtual Common::Input::NfcState StopNfcPolling(
+        [[maybe_unused]] const PadIdentifier& identifier_) {
+        return Common::Input::NfcState::NotSupported;
+    }
+    // Reads data from amiibo tag
+    virtual Common::Input::NfcState ReadAmiiboData(
+        [[maybe_unused]] const PadIdentifier& identifier_,
+        [[maybe_unused]] std::vector<u8>& out_data) {
+        return Common::Input::NfcState::NotSupported;
+    }
    // Writes data to an nfc tag
    virtual Common::Input::NfcState WriteNfcData([[maybe_unused]] const PadIdentifier& identifier,
                                                 [[maybe_unused]] const std::vector<u8>& data) {
        return Common::Input::NfcState::NotSupported;
    }
+    // Reads data from mifare tag
+    virtual Common::Input::NfcState ReadMifareData(
+        [[maybe_unused]] const PadIdentifier& identifier_,
+        [[maybe_unused]] const Common::Input::MifareRequest& request,
+        [[maybe_unused]] Common::Input::MifareRequest& out_data) {
+        return Common::Input::NfcState::NotSupported;
+    }
+    // Write data to mifare tag
+    virtual Common::Input::NfcState WriteMifareData(
+        [[maybe_unused]] const PadIdentifier& identifier_,
+        [[maybe_unused]] const Common::Input::MifareRequest& request) {
+        return Common::Input::NfcState::NotSupported;
+    }
    // Returns the engine name
    [[nodiscard]] const std::string& GetEngineName() const;
diff --git a/src/input_common/input_poller.cpp b/src/input_common/input_poller.cpp
index 380a01542..870e76ab0 100644
--- a/src/input_common/input_poller.cpp
+++ b/src/input_common/input_poller.cpp
@@ -792,8 +792,7 @@ public:
        const Common::Input::CallbackStatus status{
            .type = Common::Input::InputType::Nfc,
-            .nfc_status = nfc_status.state,
+            .nfc_status = nfc_status,
-            .raw_data = nfc_status.data,
        };
        TriggerOnChange(status);
@@ -836,10 +835,31 @@ public:
        return input_engine->SupportsNfc(identifier);
    }
+    Common::Input::NfcState StartNfcPolling() {
+        return input_engine->StartNfcPolling(identifier);
+    }
+    Common::Input::NfcState StopNfcPolling() {
+        return input_engine->StopNfcPolling(identifier);
+    }
+    Common::Input::NfcState ReadAmiiboData(std::vector<u8>& out_data) {
+        return input_engine->ReadAmiiboData(identifier, out_data);
+    }
    Common::Input::NfcState WriteNfcData(const std::vector<u8>& data) override {
        return input_engine->WriteNfcData(identifier, data);
    }
+    Common::Input::NfcState ReadMifareData(const Common::Input::MifareRequest& request,
+                                           Common::Input::MifareRequest& out_data) {
+        return input_engine->ReadMifareData(identifier, request, out_data);
+    }
+    Common::Input::NfcState WriteMifareData(const Common::Input::MifareRequest& request) {
+        return input_engine->WriteMifareData(identifier, request);
+    }
 private:
    const PadIdentifier identifier;
    InputEngine* input_engine;
diff --git a/src/shader_recompiler/backend/glsl/glsl_emit_context.cpp b/src/shader_recompiler/backend/glsl/glsl_emit_context.cpp
index c3c2281bb..9ff4028c2 100644
--- a/src/shader_recompiler/backend/glsl/glsl_emit_context.cpp
+++ b/src/shader_recompiler/backend/glsl/glsl_emit_context.cpp
@@ -479,7 +479,7 @@ void EmitContext::DefineGenericOutput(size_t index, u32 invocations) {
        const u32 remainder{4 - element};
        const TransformFeedbackVarying* xfb_varying{};
        const size_t xfb_varying_index{base_index + element};
-        if (xfb_varying_index < runtime_info.xfb_varyings.size()) {
+        if (xfb_varying_index < runtime_info.xfb_count) {
            xfb_varying = &runtime_info.xfb_varyings[xfb_varying_index];
            xfb_varying = xfb_varying->components > 0 ? xfb_varying : nullptr;
        }
diff --git a/src/shader_recompiler/backend/spirv/emit_spirv.cpp b/src/shader_recompiler/backend/spirv/emit_spirv.cpp
index 0f86a8004..34592a01f 100644
--- a/src/shader_recompiler/backend/spirv/emit_spirv.cpp
+++ b/src/shader_recompiler/backend/spirv/emit_spirv.cpp
@@ -387,7 +387,7 @@ void SetupSignedNanCapabilities(const Profile& profile, const IR::Program& progr
 }
 void SetupTransformFeedbackCapabilities(EmitContext& ctx, Id main_func) {
-    if (ctx.runtime_info.xfb_varyings.empty()) {
+    if (ctx.runtime_info.xfb_count == 0) {
        return;
    }
    ctx.AddCapability(spv::Capability::TransformFeedback);
diff --git a/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp b/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
index fd15f47ea..bec5db173 100644
--- a/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
+++ b/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
@@ -160,7 +160,7 @@ void DefineGenericOutput(EmitContext& ctx, size_t index, std::optional<u32> invo
        const u32 remainder{4 - element};
        const TransformFeedbackVarying* xfb_varying{};
        const size_t xfb_varying_index{base_attr_index + element};
-        if (xfb_varying_index < ctx.runtime_info.xfb_varyings.size()) {
+        if (xfb_varying_index < ctx.runtime_info.xfb_count) {
            xfb_varying = &ctx.runtime_info.xfb_varyings[xfb_varying_index];
            xfb_varying = xfb_varying->components > 0 ? xfb_varying : nullptr;
        }
diff --git a/src/shader_recompiler/runtime_info.h b/src/shader_recompiler/runtime_info.h
index 3b63c249f..619c0b138 100644
--- a/src/shader_recompiler/runtime_info.h
+++ b/src/shader_recompiler/runtime_info.h
@@ -84,7 +84,8 @@ struct RuntimeInfo {
    bool glasm_use_storage_buffers{};
    /// Transform feedback state for each varying
-    std::vector<TransformFeedbackVarying> xfb_varyings;
+    std::array<TransformFeedbackVarying, 256> xfb_varyings{};
+    u32 xfb_count{0};
 };
 } // namespace Shader
diff --git a/src/video_core/buffer_cache/buffer_cache.h b/src/video_core/buffer_cache/buffer_cache.h
index 45977d578..58a45ab67 100644
--- a/src/video_core/buffer_cache/buffer_cache.h
+++ b/src/video_core/buffer_cache/buffer_cache.h
@@ -207,7 +207,7 @@ bool BufferCache<P>::DMACopy(GPUVAddr src_address, GPUVAddr dest_address, u64 am
    if (has_new_downloads) {
        memory_tracker.MarkRegionAsGpuModified(*cpu_dest_address, amount);
    }
-    tmp_buffer.resize(amount);
+    tmp_buffer.resize_destructive(amount);
    cpu_memory.ReadBlockUnsafe(*cpu_src_address, tmp_buffer.data(), amount);
    cpu_memory.WriteBlockUnsafe(*cpu_dest_address, tmp_buffer.data(), amount);
    return true;
@@ -1279,7 +1279,7 @@ template <class P>
 typename BufferCache<P>::OverlapResult BufferCache<P>::ResolveOverlaps(VAddr cpu_addr,
                                                                       u32 wanted_size) {
    static constexpr int STREAM_LEAP_THRESHOLD = 16;
-    std::vector<BufferId> overlap_ids;
+    boost::container::small_vector<BufferId, 16> overlap_ids;
    VAddr begin = cpu_addr;
    VAddr end = cpu_addr + wanted_size;
    int stream_score = 0;
diff --git a/src/video_core/buffer_cache/buffer_cache_base.h b/src/video_core/buffer_cache/buffer_cache_base.h
index 63a120f7a..fe6068cfe 100644
--- a/src/video_core/buffer_cache/buffer_cache_base.h
+++ b/src/video_core/buffer_cache/buffer_cache_base.h
@@ -229,7 +229,7 @@ class BufferCache : public VideoCommon::ChannelSetupCaches<BufferCacheChannelInf
    using OverlapCounter = boost::icl::split_interval_map<VAddr, int>;
    struct OverlapResult {
-        std::vector<BufferId> ids;
+        boost::container::small_vector<BufferId, 16> ids;
        VAddr begin;
        VAddr end;
        bool has_stream_leap = false;
@@ -582,7 +582,7 @@ private:
    BufferId inline_buffer_id;
    std::array<BufferId, ((1ULL << 39) >> CACHING_PAGEBITS)> page_table;
-    std::vector<u8> tmp_buffer;
+    Common::ScratchBuffer<u8> tmp_buffer;
 };
 } // namespace VideoCommon
diff --git a/src/video_core/cdma_pusher.h b/src/video_core/cdma_pusher.h
index 83112dfce..7d660af47 100644
--- a/src/video_core/cdma_pusher.h
+++ b/src/video_core/cdma_pusher.h
@@ -63,7 +63,6 @@ struct ChCommand {
 };
 using ChCommandHeaderList = std::vector<ChCommandHeader>;
-using ChCommandList = std::vector<ChCommand>;
 struct ThiRegisters {
    u32_le increment_syncpt{};
diff --git a/src/video_core/dma_pusher.h b/src/video_core/dma_pusher.h
index 1cdb690ed..8a2784cdc 100644
--- a/src/video_core/dma_pusher.h
+++ b/src/video_core/dma_pusher.h
@@ -6,6 +6,7 @@
 #include <array>
 #include <span>
 #include <vector>
+#include <boost/container/small_vector.hpp>
 #include <queue>
 #include "common/bit_field.h"
@@ -102,11 +103,12 @@ inline CommandHeader BuildCommandHeader(BufferMethods method, u32 arg_count, Sub
 struct CommandList final {
    CommandList() = default;
    explicit CommandList(std::size_t size) : command_lists(size) {}
-    explicit CommandList(std::vector<CommandHeader>&& prefetch_command_list_)
+    explicit CommandList(
+        boost::container::small_vector<CommandHeader, 512>&& prefetch_command_list_)
        : prefetch_command_list{std::move(prefetch_command_list_)} {}
-    std::vector<CommandListHeader> command_lists;
+    boost::container::small_vector<CommandListHeader, 512> command_lists;
-    std::vector<CommandHeader> prefetch_command_list;
+    boost::container::small_vector<CommandHeader, 512> prefetch_command_list;
 };
 /**
diff --git a/src/video_core/engines/maxwell_dma.cpp b/src/video_core/engines/maxwell_dma.cpp
index ebe5536de..bc1eb41e7 100644
--- a/src/video_core/engines/maxwell_dma.cpp
+++ b/src/video_core/engines/maxwell_dma.cpp
@@ -108,9 +108,11 @@ void MaxwellDMA::Launch() {
        if (regs.launch_dma.remap_enable != 0 && is_const_a_dst) {
            ASSERT(regs.remap_const.component_size_minus_one == 3);
            accelerate.BufferClear(regs.offset_out, regs.line_length_in, regs.remap_consta_value);
-            std::vector<u32> tmp_buffer(regs.line_length_in, regs.remap_consta_value);
+            read_buffer.resize_destructive(regs.line_length_in * sizeof(u32));
+            std::span<u32> span(reinterpret_cast<u32*>(read_buffer.data()), regs.line_length_in);
+            std::ranges::fill(span, regs.remap_consta_value);
            memory_manager.WriteBlockUnsafe(regs.offset_out,
-                                            reinterpret_cast<u8*>(tmp_buffer.data()),
+                                            reinterpret_cast<u8*>(read_buffer.data()),
                                            regs.line_length_in * sizeof(u32));
        } else {
            memory_manager.FlushCaching();
@@ -126,32 +128,32 @@ void MaxwellDMA::Launch() {
                UNIMPLEMENTED_IF(regs.line_length_in % 16 != 0);
                UNIMPLEMENTED_IF(regs.offset_in % 16 != 0);
                UNIMPLEMENTED_IF(regs.offset_out % 16 != 0);
-                std::vector<u8> tmp_buffer(16);
+                read_buffer.resize_destructive(16);
                for (u32 offset = 0; offset < regs.line_length_in; offset += 16) {
                    memory_manager.ReadBlockUnsafe(
                        convert_linear_2_blocklinear_addr(regs.offset_in + offset),
-                        tmp_buffer.data(), tmp_buffer.size());
+                        read_buffer.data(), read_buffer.size());
-                    memory_manager.WriteBlockCached(regs.offset_out + offset, tmp_buffer.data(),
+                    memory_manager.WriteBlockCached(regs.offset_out + offset, read_buffer.data(),
-                                                    tmp_buffer.size());
+                                                    read_buffer.size());
                }
            } else if (is_src_pitch && !is_dst_pitch) {
                UNIMPLEMENTED_IF(regs.line_length_in % 16 != 0);
                UNIMPLEMENTED_IF(regs.offset_in % 16 != 0);
                UNIMPLEMENTED_IF(regs.offset_out % 16 != 0);
-                std::vector<u8> tmp_buffer(16);
+                read_buffer.resize_destructive(16);
                for (u32 offset = 0; offset < regs.line_length_in; offset += 16) {
-                    memory_manager.ReadBlockUnsafe(regs.offset_in + offset, tmp_buffer.data(),
+                    memory_manager.ReadBlockUnsafe(regs.offset_in + offset, read_buffer.data(),
-                                                   tmp_buffer.size());
+                                                   read_buffer.size());
                    memory_manager.WriteBlockCached(
                        convert_linear_2_blocklinear_addr(regs.offset_out + offset),
-                        tmp_buffer.data(), tmp_buffer.size());
+                        read_buffer.data(), read_buffer.size());
                }
            } else {
                if (!accelerate.BufferCopy(regs.offset_in, regs.offset_out, regs.line_length_in)) {
-                    std::vector<u8> tmp_buffer(regs.line_length_in);
+                    read_buffer.resize_destructive(regs.line_length_in);
-                    memory_manager.ReadBlockUnsafe(regs.offset_in, tmp_buffer.data(),
+                    memory_manager.ReadBlockUnsafe(regs.offset_in, read_buffer.data(),
                                                   regs.line_length_in);
-                    memory_manager.WriteBlockCached(regs.offset_out, tmp_buffer.data(),
+                    memory_manager.WriteBlockCached(regs.offset_out, read_buffer.data(),
                                                    regs.line_length_in);
                }
            }
@@ -171,7 +173,8 @@ void MaxwellDMA::CopyBlockLinearToPitch() {
    src_operand.address = regs.offset_in;
    DMA::BufferOperand dst_operand;
-    dst_operand.pitch = regs.pitch_out;
+    u32 abs_pitch_out = std::abs(static_cast<s32>(regs.pitch_out));
+    dst_operand.pitch = abs_pitch_out;
    dst_operand.width = regs.line_length_in;
    dst_operand.height = regs.line_count;
    dst_operand.address = regs.offset_out;
@@ -218,7 +221,7 @@ void MaxwellDMA::CopyBlockLinearToPitch() {
    const size_t src_size =
        CalculateSize(true, bytes_per_pixel, width, height, depth, block_height, block_depth);
-    const size_t dst_size = static_cast<size_t>(regs.pitch_out) * regs.line_count;
+    const size_t dst_size = static_cast<size_t>(abs_pitch_out) * regs.line_count;
    read_buffer.resize_destructive(src_size);
    write_buffer.resize_destructive(dst_size);
@@ -227,7 +230,7 @@ void MaxwellDMA::CopyBlockLinearToPitch() {
    UnswizzleSubrect(write_buffer, read_buffer, bytes_per_pixel, width, height, depth, x_offset,
                     src_params.origin.y, x_elements, regs.line_count, block_height, block_depth,
-                     regs.pitch_out);
+                     abs_pitch_out);
    memory_manager.WriteBlockCached(regs.offset_out, write_buffer.data(), dst_size);
 }
diff --git a/src/video_core/gpu.cpp b/src/video_core/gpu.cpp
index 456f733cf..db385076d 100644
--- a/src/video_core/gpu.cpp
+++ b/src/video_core/gpu.cpp
@@ -193,18 +193,13 @@ struct GPU::Impl {
    }
    [[nodiscard]] u64 GetTicks() const {
-        // This values were reversed engineered by fincs from NVN
+        u64 gpu_tick = system.CoreTiming().GetGPUTicks();
-        // The gpu clock is reported in units of 385/625 nanoseconds
-        constexpr u64 gpu_ticks_num = 384;
-        constexpr u64 gpu_ticks_den = 625;
-        u64 nanoseconds = system.CoreTiming().GetCPUTimeNs().count();
        if (Settings::values.use_fast_gpu_time.GetValue()) {
-            nanoseconds /= 256;
+            gpu_tick /= 256;
        }
-        const u64 nanoseconds_num = nanoseconds / gpu_ticks_den;
-        const u64 nanoseconds_rem = nanoseconds % gpu_ticks_den;
+        return gpu_tick;
-        return nanoseconds_num * gpu_ticks_num + (nanoseconds_rem * gpu_ticks_num) / gpu_ticks_den;
    }
    [[nodiscard]] bool IsAsync() const {
diff --git a/src/video_core/host1x/codecs/h264.cpp b/src/video_core/host1x/codecs/h264.cpp
index 6ce179167..ce827eb6c 100644
--- a/src/video_core/host1x/codecs/h264.cpp
+++ b/src/video_core/host1x/codecs/h264.cpp
@@ -4,6 +4,7 @@
 #include <array>
 #include <bit>
+#include "common/scratch_buffer.h"
 #include "common/settings.h"
 #include "video_core/host1x/codecs/h264.h"
 #include "video_core/host1x/host1x.h"
@@ -188,7 +189,8 @@ void H264BitWriter::WriteBit(bool state) {
 }
 void H264BitWriter::WriteScalingList(std::span<const u8> list, s32 start, s32 count) {
-    std::vector<u8> scan(count);
+    static Common::ScratchBuffer<u8> scan{};
+    scan.resize_destructive(count);
    if (count == 16) {
        std::memcpy(scan.data(), zig_zag_scan.data(), scan.size());
    } else {
diff --git a/src/video_core/memory_manager.cpp b/src/video_core/memory_manager.cpp
index b2f7e160a..45141e488 100644
--- a/src/video_core/memory_manager.cpp
+++ b/src/video_core/memory_manager.cpp
@@ -587,7 +587,7 @@ void MemoryManager::InvalidateRegion(GPUVAddr gpu_addr, size_t size,
 void MemoryManager::CopyBlock(GPUVAddr gpu_dest_addr, GPUVAddr gpu_src_addr, std::size_t size,
                              VideoCommon::CacheType which) {
-    std::vector<u8> tmp_buffer(size);
+    tmp_buffer.resize_destructive(size);
    ReadBlock(gpu_src_addr, tmp_buffer.data(), size, which);
    // The output block must be flushed in case it has data modified from the GPU.
@@ -670,9 +670,9 @@ bool MemoryManager::IsFullyMappedRange(GPUVAddr gpu_addr, std::size_t size) cons
    return result;
 }
-std::vector<std::pair<GPUVAddr, std::size_t>> MemoryManager::GetSubmappedRange(
+boost::container::small_vector<std::pair<GPUVAddr, std::size_t>, 32>
-    GPUVAddr gpu_addr, std::size_t size) const {
+MemoryManager::GetSubmappedRange(GPUVAddr gpu_addr, std::size_t size) const {
-    std::vector<std::pair<GPUVAddr, std::size_t>> result{};
+    boost::container::small_vector<std::pair<GPUVAddr, std::size_t>, 32> result{};
    GetSubmappedRangeImpl<true>(gpu_addr, size, result);
    return result;
 }
@@ -680,8 +680,9 @@ std::vector<std::pair<GPUVAddr, std::size_t>> MemoryManager::GetSubmappedRange(
 template <bool is_gpu_address>
 void MemoryManager::GetSubmappedRangeImpl(
    GPUVAddr gpu_addr, std::size_t size,
-    std::vector<std::pair<std::conditional_t<is_gpu_address, GPUVAddr, VAddr>, std::size_t>>&
+    boost::container::small_vector<
-        result) const {
+        std::pair<std::conditional_t<is_gpu_address, GPUVAddr, VAddr>, std::size_t>, 32>& result)
+    const {
    std::optional<std::pair<std::conditional_t<is_gpu_address, GPUVAddr, VAddr>, std::size_t>>
        last_segment{};
    std::optional<VAddr> old_page_addr{};
diff --git a/src/video_core/memory_manager.h b/src/video_core/memory_manager.h
index 794535122..4202c26ff 100644
--- a/src/video_core/memory_manager.h
+++ b/src/video_core/memory_manager.h
@@ -8,10 +8,12 @@
 #include <mutex>
 #include <optional>
 #include <vector>
+#include <boost/container/small_vector.hpp>
 #include "common/common_types.h"
 #include "common/multi_level_page_table.h"
 #include "common/range_map.h"
+#include "common/scratch_buffer.h"
 #include "common/virtual_buffer.h"
 #include "video_core/cache_types.h"
 #include "video_core/pte_kind.h"
@@ -107,8 +109,8 @@ public:
     * if the region is continuous, a single pair will be returned. If it's unmapped, an empty
     * vector will be returned;
     */
-    std::vector<std::pair<GPUVAddr, std::size_t>> GetSubmappedRange(GPUVAddr gpu_addr,
+    boost::container::small_vector<std::pair<GPUVAddr, std::size_t>, 32> GetSubmappedRange(
-                                                                    std::size_t size) const;
+        GPUVAddr gpu_addr, std::size_t size) const;
    GPUVAddr Map(GPUVAddr gpu_addr, VAddr cpu_addr, std::size_t size,
                 PTEKind kind = PTEKind::INVALID, bool is_big_pages = true);
@@ -165,7 +167,8 @@ private:
    template <bool is_gpu_address>
    void GetSubmappedRangeImpl(
        GPUVAddr gpu_addr, std::size_t size,
-        std::vector<std::pair<std::conditional_t<is_gpu_address, GPUVAddr, VAddr>, std::size_t>>&
+        boost::container::small_vector<
+            std::pair<std::conditional_t<is_gpu_address, GPUVAddr, VAddr>, std::size_t>, 32>&
            result) const;
    Core::System& system;
@@ -215,8 +218,8 @@ private:
    Common::VirtualBuffer<u32> big_page_table_cpu;
    std::vector<u64> big_page_continuous;
-    std::vector<std::pair<VAddr, std::size_t>> page_stash{};
+    boost::container::small_vector<std::pair<VAddr, std::size_t>, 32> page_stash{};
-    std::vector<std::pair<VAddr, std::size_t>> page_stash2{};
+    boost::container::small_vector<std::pair<VAddr, std::size_t>, 32> page_stash2{};
    mutable std::mutex guard;
@@ -226,6 +229,8 @@ private:
    std::unique_ptr<VideoCommon::InvalidationAccumulator> accumulator;
    static std::atomic<size_t> unique_identifier_generator;
+    Common::ScratchBuffer<u8> tmp_buffer;
 };
 } // namespace Tegra
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.cpp b/src/video_core/renderer_opengl/gl_shader_cache.cpp
index 3f077311e..0329ed820 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_cache.cpp
@@ -85,7 +85,9 @@ Shader::RuntimeInfo MakeRuntimeInfo(const GraphicsPipelineKey& key,
    case Shader::Stage::VertexB:
    case Shader::Stage::Geometry:
        if (!use_assembly_shaders && key.xfb_enabled != 0) {
-            info.xfb_varyings = VideoCommon::MakeTransformFeedbackVaryings(key.xfb_state);
+            auto [varyings, count] = VideoCommon::MakeTransformFeedbackVaryings(key.xfb_state);
+            info.xfb_varyings = varyings;
+            info.xfb_count = count;
        }
        break;
    case Shader::Stage::TessellationEval:
diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
index e30fcb1ed..f47301ad5 100644
--- a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
+++ b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
@@ -361,7 +361,7 @@ void BufferCacheRuntime::CopyBuffer(VkBuffer dst_buffer, VkBuffer src_buffer,
        .dstAccessMask = VK_ACCESS_MEMORY_READ_BIT | VK_ACCESS_MEMORY_WRITE_BIT,
    };
    // Measuring a popular game, this number never exceeds the specified size once data is warmed up
-    boost::container::small_vector<VkBufferCopy, 3> vk_copies(copies.size());
+    boost::container::small_vector<VkBufferCopy, 8> vk_copies(copies.size());
    std::ranges::transform(copies, vk_copies.begin(), MakeBufferCopy);
    scheduler.RequestOutsideRenderPassOperationContext();
    scheduler.Record([src_buffer, dst_buffer, vk_copies, barrier](vk::CommandBuffer cmdbuf) {
diff --git a/src/video_core/renderer_vulkan/vk_master_semaphore.cpp b/src/video_core/renderer_vulkan/vk_master_semaphore.cpp
index 5eeda08d2..6b288b994 100644
--- a/src/video_core/renderer_vulkan/vk_master_semaphore.cpp
+++ b/src/video_core/renderer_vulkan/vk_master_semaphore.cpp
@@ -75,15 +75,9 @@ void MasterSemaphore::Refresh() {
 void MasterSemaphore::Wait(u64 tick) {
    if (!semaphore) {
-        // If we don't support timeline semaphores, use an atomic wait
+        // If we don't support timeline semaphores, wait for the value normally
-        while (true) {
+        std::unique_lock lk{free_mutex};
-            u64 current_value = gpu_tick.load(std::memory_order_relaxed);
+        free_cv.wait(lk, [&] { return gpu_tick.load(std::memory_order_relaxed) >= tick; });
-            if (current_value >= tick) {
-                return;
-            }
-            gpu_tick.wait(current_value);
-        }
        return;
    }
@@ -198,11 +192,13 @@ void MasterSemaphore::WaitThread(std::stop_token token) {
        fence.Wait();
        fence.Reset();
-        gpu_tick.store(host_tick);
-        gpu_tick.notify_all();
-        std::scoped_lock lock{free_mutex};
+        {
-        free_queue.push_front(std::move(fence));
+            std::scoped_lock lock{free_mutex};
+            free_queue.push_front(std::move(fence));
+            gpu_tick.store(host_tick);
+        }
+        free_cv.notify_one();
    }
 }
diff --git a/src/video_core/renderer_vulkan/vk_master_semaphore.h b/src/video_core/renderer_vulkan/vk_master_semaphore.h
index 1e7c90215..3f599d7bd 100644
--- a/src/video_core/renderer_vulkan/vk_master_semaphore.h
+++ b/src/video_core/renderer_vulkan/vk_master_semaphore.h
@@ -72,6 +72,7 @@ private:
    std::atomic<u64> current_tick{1}; ///< Current logical tick.
    std::mutex wait_mutex;
    std::mutex free_mutex;
+    std::condition_variable free_cv;
    std::condition_variable_any wait_cv;
    std::queue<Waitable> wait_queue;  ///< Queue for the fences to be waited on by the wait thread.
    std::deque<vk::Fence> free_queue; ///< Holds available fences for submission.
diff --git a/src/video_core/renderer_vulkan/vk_pipeline_cache.cpp b/src/video_core/renderer_vulkan/vk_pipeline_cache.cpp
index a2cfb2105..9f316113c 100644
--- a/src/video_core/renderer_vulkan/vk_pipeline_cache.cpp
+++ b/src/video_core/renderer_vulkan/vk_pipeline_cache.cpp
@@ -167,7 +167,10 @@ Shader::RuntimeInfo MakeRuntimeInfo(std::span<const Shader::IR::Program> program
                info.fixed_state_point_size = point_size;
            }
            if (key.state.xfb_enabled) {
-                info.xfb_varyings = VideoCommon::MakeTransformFeedbackVaryings(key.state.xfb_state);
+                auto [varyings, count] =
+                    VideoCommon::MakeTransformFeedbackVaryings(key.state.xfb_state);
+                info.xfb_varyings = varyings;
+                info.xfb_count = count;
            }
            info.convert_depth_mode = gl_ndc;
        }
@@ -214,7 +217,10 @@ Shader::RuntimeInfo MakeRuntimeInfo(std::span<const Shader::IR::Program> program
            info.fixed_state_point_size = point_size;
        }
        if (key.state.xfb_enabled != 0) {
-            info.xfb_varyings = VideoCommon::MakeTransformFeedbackVaryings(key.state.xfb_state);
+            auto [varyings, count] =
+                VideoCommon::MakeTransformFeedbackVaryings(key.state.xfb_state);
+            info.xfb_varyings = varyings;
+            info.xfb_count = count;
        }
        info.convert_depth_mode = gl_ndc;
        break;
diff --git a/src/video_core/renderer_vulkan/vk_texture_cache.cpp b/src/video_core/renderer_vulkan/vk_texture_cache.cpp
index f025f618b..f3cef09dd 100644
--- a/src/video_core/renderer_vulkan/vk_texture_cache.cpp
+++ b/src/video_core/renderer_vulkan/vk_texture_cache.cpp
@@ -330,9 +330,9 @@ constexpr VkBorderColor ConvertBorderColor(const std::array<float, 4>& color) {
    };
 }
-[[maybe_unused]] [[nodiscard]] std::vector<VkBufferCopy> TransformBufferCopies(
+[[maybe_unused]] [[nodiscard]] boost::container::small_vector<VkBufferCopy, 16>
-    std::span<const VideoCommon::BufferCopy> copies, size_t buffer_offset) {
+TransformBufferCopies(std::span<const VideoCommon::BufferCopy> copies, size_t buffer_offset) {
-    std::vector<VkBufferCopy> result(copies.size());
+    boost::container::small_vector<VkBufferCopy, 16> result(copies.size());
    std::ranges::transform(
        copies, result.begin(), [buffer_offset](const VideoCommon::BufferCopy& copy) {
            return VkBufferCopy{
@@ -344,7 +344,7 @@ constexpr VkBorderColor ConvertBorderColor(const std::array<float, 4>& color) {
    return result;
 }
-[[nodiscard]] std::vector<VkBufferImageCopy> TransformBufferImageCopies(
+[[nodiscard]] boost::container::small_vector<VkBufferImageCopy, 16> TransformBufferImageCopies(
    std::span<const BufferImageCopy> copies, size_t buffer_offset, VkImageAspectFlags aspect_mask) {
    struct Maker {
        VkBufferImageCopy operator()(const BufferImageCopy& copy) const {
@@ -377,14 +377,14 @@ constexpr VkBorderColor ConvertBorderColor(const std::array<float, 4>& color) {
        VkImageAspectFlags aspect_mask;
    };
    if (aspect_mask == (VK_IMAGE_ASPECT_DEPTH_BIT | VK_IMAGE_ASPECT_STENCIL_BIT)) {
-        std::vector<VkBufferImageCopy> result(copies.size() * 2);
+        boost::container::small_vector<VkBufferImageCopy, 16> result(copies.size() * 2);
        std::ranges::transform(copies, result.begin(),
                               Maker{buffer_offset, VK_IMAGE_ASPECT_DEPTH_BIT});
        std::ranges::transform(copies, result.begin() + copies.size(),
                               Maker{buffer_offset, VK_IMAGE_ASPECT_STENCIL_BIT});
        return result;
    } else {
-        std::vector<VkBufferImageCopy> result(copies.size());
+        boost::container::small_vector<VkBufferImageCopy, 16> result(copies.size());
        std::ranges::transform(copies, result.begin(), Maker{buffer_offset, aspect_mask});
        return result;
    }
@@ -867,8 +867,8 @@ void TextureCacheRuntime::BarrierFeedbackLoop() {
 void TextureCacheRuntime::ReinterpretImage(Image& dst, Image& src,
                                           std::span<const VideoCommon::ImageCopy> copies) {
-    std::vector<VkBufferImageCopy> vk_in_copies(copies.size());
+    boost::container::small_vector<VkBufferImageCopy, 16> vk_in_copies(copies.size());
-    std::vector<VkBufferImageCopy> vk_out_copies(copies.size());
+    boost::container::small_vector<VkBufferImageCopy, 16> vk_out_copies(copies.size());
    const VkImageAspectFlags src_aspect_mask = src.AspectMask();
    const VkImageAspectFlags dst_aspect_mask = dst.AspectMask();
@@ -1157,7 +1157,7 @@ void TextureCacheRuntime::ConvertImage(Framebuffer* dst, ImageView& dst_view, Im
 void TextureCacheRuntime::CopyImage(Image& dst, Image& src,
                                    std::span<const VideoCommon::ImageCopy> copies) {
-    std::vector<VkImageCopy> vk_copies(copies.size());
+    boost::container::small_vector<VkImageCopy, 16> vk_copies(copies.size());
    const VkImageAspectFlags aspect_mask = dst.AspectMask();
    ASSERT(aspect_mask == src.AspectMask());
@@ -1332,7 +1332,7 @@ void Image::UploadMemory(VkBuffer buffer, VkDeviceSize offset,
        ScaleDown(true);
    }
    scheduler->RequestOutsideRenderPassOperationContext();
-    std::vector vk_copies = TransformBufferImageCopies(copies, offset, aspect_mask);
+    auto vk_copies = TransformBufferImageCopies(copies, offset, aspect_mask);
    const VkBuffer src_buffer = buffer;
    const VkImage vk_image = *original_image;
    const VkImageAspectFlags vk_aspect_mask = aspect_mask;
@@ -1367,8 +1367,9 @@ void Image::DownloadMemory(std::span<VkBuffer> buffers_span, std::span<VkDeviceS
    if (is_rescaled) {
        ScaleDown();
    }
-    boost::container::small_vector<VkBuffer, 1> buffers_vector{};
+    boost::container::small_vector<VkBuffer, 8> buffers_vector{};
-    boost::container::small_vector<std::vector<VkBufferImageCopy>, 1> vk_copies;
+    boost::container::small_vector<boost::container::small_vector<VkBufferImageCopy, 16>, 8>
+        vk_copies;
    for (size_t index = 0; index < buffers_span.size(); index++) {
        buffers_vector.emplace_back(buffers_span[index]);
        vk_copies.emplace_back(
@@ -1858,7 +1859,7 @@ Framebuffer::~Framebuffer() = default;
 void Framebuffer::CreateFramebuffer(TextureCacheRuntime& runtime,
                                    std::span<ImageView*, NUM_RT> color_buffers,
                                    ImageView* depth_buffer, bool is_rescaled) {
-    std::vector<VkImageView> attachments;
+    boost::container::small_vector<VkImageView, NUM_RT + 1> attachments;
    RenderPassKey renderpass_key{};
    s32 num_layers = 1;
diff --git a/src/video_core/shader_cache.cpp b/src/video_core/shader_cache.cpp
index c5213875b..4db948b6d 100644
--- a/src/video_core/shader_cache.cpp
+++ b/src/video_core/shader_cache.cpp
@@ -151,11 +151,9 @@ void ShaderCache::RemovePendingShaders() {
    marked_for_removal.erase(std::unique(marked_for_removal.begin(), marked_for_removal.end()),
                             marked_for_removal.end());
-    std::vector<ShaderInfo*> removed_shaders;
+    boost::container::small_vector<ShaderInfo*, 16> removed_shaders;
-    removed_shaders.reserve(marked_for_removal.size());
    std::scoped_lock lock{lookup_mutex};
    for (Entry* const entry : marked_for_removal) {
        removed_shaders.push_back(entry->data);
diff --git a/src/video_core/texture_cache/image_base.h b/src/video_core/texture_cache/image_base.h
index 1b8a17ee8..55d49d017 100644
--- a/src/video_core/texture_cache/image_base.h
+++ b/src/video_core/texture_cache/image_base.h
@@ -6,6 +6,7 @@
 #include <array>
 #include <optional>
 #include <vector>
+#include <boost/container/small_vector.hpp>
 #include "common/common_funcs.h"
 #include "common/common_types.h"
@@ -108,8 +109,8 @@ struct ImageBase {
    std::vector<ImageViewInfo> image_view_infos;
    std::vector<ImageViewId> image_view_ids;
-    std::vector<u32> slice_offsets;
+    boost::container::small_vector<u32, 16> slice_offsets;
-    std::vector<SubresourceBase> slice_subresources;
+    boost::container::small_vector<SubresourceBase, 16> slice_subresources;
    std::vector<AliasedImage> aliased_images;
    std::vector<ImageId> overlapping_images;
diff --git a/src/video_core/texture_cache/texture_cache.h b/src/video_core/texture_cache/texture_cache.h
index d25339c8c..d3f03a995 100644
--- a/src/video_core/texture_cache/texture_cache.h
+++ b/src/video_core/texture_cache/texture_cache.h
@@ -186,6 +186,10 @@ void TextureCache<P>::FillComputeImageViews(std::span<ImageViewInOut> views) {
 template <class P>
 void TextureCache<P>::CheckFeedbackLoop(std::span<const ImageViewInOut> views) {
+    if (!Settings::values.barrier_feedback_loops.GetValue()) {
+        return;
+    }
    const bool requires_barrier = [&] {
        for (const auto& view : views) {
            if (!view.id) {
@@ -522,7 +526,7 @@ void TextureCache<P>::WriteMemory(VAddr cpu_addr, size_t size) {
 template <class P>
 void TextureCache<P>::DownloadMemory(VAddr cpu_addr, size_t size) {
-    std::vector<ImageId> images;
+    boost::container::small_vector<ImageId, 16> images;
    ForEachImageInRegion(cpu_addr, size, [&images](ImageId image_id, ImageBase& image) {
        if (!image.IsSafeDownload()) {
            return;
@@ -575,7 +579,7 @@ std::optional<VideoCore::RasterizerDownloadArea> TextureCache<P>::GetFlushArea(V
 template <class P>
 void TextureCache<P>::UnmapMemory(VAddr cpu_addr, size_t size) {
-    std::vector<ImageId> deleted_images;
+    boost::container::small_vector<ImageId, 16> deleted_images;
    ForEachImageInRegion(cpu_addr, size, [&](ImageId id, Image&) { deleted_images.push_back(id); });
    for (const ImageId id : deleted_images) {
        Image& image = slot_images[id];
@@ -589,7 +593,7 @@ void TextureCache<P>::UnmapMemory(VAddr cpu_addr, size_t size) {
 template <class P>
 void TextureCache<P>::UnmapGPUMemory(size_t as_id, GPUVAddr gpu_addr, size_t size) {
-    std::vector<ImageId> deleted_images;
+    boost::container::small_vector<ImageId, 16> deleted_images;
    ForEachImageInRegionGPU(as_id, gpu_addr, size,
                            [&](ImageId id, Image&) { deleted_images.push_back(id); });
    for (const ImageId id : deleted_images) {
@@ -1097,7 +1101,7 @@ ImageId TextureCache<P>::FindImage(const ImageInfo& info, GPUVAddr gpu_addr,
    const bool native_bgr = runtime.HasNativeBgr();
    const bool flexible_formats = True(options & RelaxedOptions::Format);
    ImageId image_id{};
-    boost::container::small_vector<ImageId, 1> image_ids;
+    boost::container::small_vector<ImageId, 8> image_ids;
    const auto lambda = [&](ImageId existing_image_id, ImageBase& existing_image) {
        if (True(existing_image.flags & ImageFlagBits::Remapped)) {
            return false;
@@ -1618,7 +1622,7 @@ ImageId TextureCache<P>::FindDMAImage(const ImageInfo& info, GPUVAddr gpu_addr)
        }
    }
    ImageId image_id{};
-    boost::container::small_vector<ImageId, 1> image_ids;
+    boost::container::small_vector<ImageId, 8> image_ids;
    const auto lambda = [&](ImageId existing_image_id, ImageBase& existing_image) {
        if (True(existing_image.flags & ImageFlagBits::Remapped)) {
            return false;
@@ -1938,7 +1942,7 @@ void TextureCache<P>::RegisterImage(ImageId image_id) {
        image.map_view_id = map_id;
        return;
    }
-    std::vector<ImageViewId> sparse_maps{};
+    boost::container::small_vector<ImageViewId, 16> sparse_maps;
    ForEachSparseSegment(
        image, [this, image_id, &sparse_maps](GPUVAddr gpu_addr, VAddr cpu_addr, size_t size) {
            auto map_id = slot_map_views.insert(gpu_addr, cpu_addr, size, image_id);
@@ -2213,7 +2217,7 @@ void TextureCache<P>::MarkModification(ImageBase& image) noexcept {
 template <class P>
 void TextureCache<P>::SynchronizeAliases(ImageId image_id) {
-    boost::container::small_vector<const AliasedImage*, 1> aliased_images;
+    boost::container::small_vector<const AliasedImage*, 8> aliased_images;
    Image& image = slot_images[image_id];
    bool any_rescaled = True(image.flags & ImageFlagBits::Rescaled);
    bool any_modified = True(image.flags & ImageFlagBits::GpuModified);
diff --git a/src/video_core/texture_cache/texture_cache_base.h b/src/video_core/texture_cache/texture_cache_base.h
index 44232b961..e9ec91265 100644
--- a/src/video_core/texture_cache/texture_cache_base.h
+++ b/src/video_core/texture_cache/texture_cache_base.h
@@ -56,7 +56,7 @@ struct ImageViewInOut {
 struct AsyncDecodeContext {
    ImageId image_id;
    Common::ScratchBuffer<u8> decoded_data;
-    std::vector<BufferImageCopy> copies;
+    boost::container::small_vector<BufferImageCopy, 16> copies;
    std::mutex mutex;
    std::atomic_bool complete;
 };
@@ -429,7 +429,7 @@ private:
    std::unordered_map<u64, std::vector<ImageMapId>, Common::IdentityHash<u64>> page_table;
    std::unordered_map<u64, std::vector<ImageId>, Common::IdentityHash<u64>> sparse_page_table;
-    std::unordered_map<ImageId, std::vector<ImageViewId>> sparse_views;
+    std::unordered_map<ImageId, boost::container::small_vector<ImageViewId, 16>> sparse_views;
    VAddr virtual_invalid_space{};
diff --git a/src/video_core/texture_cache/util.cpp b/src/video_core/texture_cache/util.cpp
index 95a5b47d8..f781cb7a0 100644
--- a/src/video_core/texture_cache/util.cpp
+++ b/src/video_core/texture_cache/util.cpp
@@ -329,13 +329,13 @@ template <u32 GOB_EXTENT>
 [[nodiscard]] std::optional<SubresourceExtent> ResolveOverlapRightAddress3D(
    const ImageInfo& new_info, GPUVAddr gpu_addr, const ImageBase& overlap, bool strict_size) {
-    const std::vector<u32> slice_offsets = CalculateSliceOffsets(new_info);
+    const auto slice_offsets = CalculateSliceOffsets(new_info);
    const u32 diff = static_cast<u32>(overlap.gpu_addr - gpu_addr);
    const auto it = std::ranges::find(slice_offsets, diff);
    if (it == slice_offsets.end()) {
        return std::nullopt;
    }
-    const std::vector subresources = CalculateSliceSubresources(new_info);
+    const auto subresources = CalculateSliceSubresources(new_info);
    const SubresourceBase base = subresources[std::distance(slice_offsets.begin(), it)];
    const ImageInfo& info = overlap.info;
    if (!IsBlockLinearSizeCompatible(new_info, info, base.level, 0, strict_size)) {
@@ -655,9 +655,9 @@ LevelArray CalculateMipLevelSizes(const ImageInfo& info) noexcept {
    return sizes;
 }
-std::vector<u32> CalculateSliceOffsets(const ImageInfo& info) {
+boost::container::small_vector<u32, 16> CalculateSliceOffsets(const ImageInfo& info) {
    ASSERT(info.type == ImageType::e3D);
-    std::vector<u32> offsets;
+    boost::container::small_vector<u32, 16> offsets;
    offsets.reserve(NumSlices(info));
    const LevelInfo level_info = MakeLevelInfo(info);
@@ -679,9 +679,10 @@ std::vector<u32> CalculateSliceOffsets(const ImageInfo& info) {
    return offsets;
 }
-std::vector<SubresourceBase> CalculateSliceSubresources(const ImageInfo& info) {
+boost::container::small_vector<SubresourceBase, 16> CalculateSliceSubresources(
+    const ImageInfo& info) {
    ASSERT(info.type == ImageType::e3D);
-    std::vector<SubresourceBase> subresources;
+    boost::container::small_vector<SubresourceBase, 16> subresources;
    subresources.reserve(NumSlices(info));
    for (s32 level = 0; level < info.resources.levels; ++level) {
        const s32 depth = AdjustMipSize(info.size.depth, level);
@@ -723,8 +724,10 @@ ImageViewType RenderTargetImageViewType(const ImageInfo& info) noexcept {
    }
 }
-std::vector<ImageCopy> MakeShrinkImageCopies(const ImageInfo& dst, const ImageInfo& src,
+boost::container::small_vector<ImageCopy, 16> MakeShrinkImageCopies(const ImageInfo& dst,
-                                             SubresourceBase base, u32 up_scale, u32 down_shift) {
+                                                                    const ImageInfo& src,
+                                                                    SubresourceBase base,
+                                                                    u32 up_scale, u32 down_shift) {
    ASSERT(dst.resources.levels >= src.resources.levels);
    const bool is_dst_3d = dst.type == ImageType::e3D;
@@ -733,7 +736,7 @@ std::vector<ImageCopy> MakeShrinkImageCopies(const ImageInfo& dst, const ImageIn
        ASSERT(src.resources.levels == 1);
    }
    const bool both_2d{src.type == ImageType::e2D && dst.type == ImageType::e2D};
-    std::vector<ImageCopy> copies;
+    boost::container::small_vector<ImageCopy, 16> copies;
    copies.reserve(src.resources.levels);
    for (s32 level = 0; level < src.resources.levels; ++level) {
        ImageCopy& copy = copies.emplace_back();
@@ -770,9 +773,10 @@ std::vector<ImageCopy> MakeShrinkImageCopies(const ImageInfo& dst, const ImageIn
    return copies;
 }
-std::vector<ImageCopy> MakeReinterpretImageCopies(const ImageInfo& src, u32 up_scale,
+boost::container::small_vector<ImageCopy, 16> MakeReinterpretImageCopies(const ImageInfo& src,
-                                                  u32 down_shift) {
+                                                                         u32 up_scale,
-    std::vector<ImageCopy> copies;
+                                                                         u32 down_shift) {
+    boost::container::small_vector<ImageCopy, 16> copies;
    copies.reserve(src.resources.levels);
    const bool is_3d = src.type == ImageType::e3D;
    for (s32 level = 0; level < src.resources.levels; ++level) {
@@ -824,9 +828,11 @@ bool IsValidEntry(const Tegra::MemoryManager& gpu_memory, const TICEntry& config
    return gpu_memory.GpuToCpuAddress(address, guest_size_bytes).has_value();
 }
-std::vector<BufferImageCopy> UnswizzleImage(Tegra::MemoryManager& gpu_memory, GPUVAddr gpu_addr,
+boost::container::small_vector<BufferImageCopy, 16> UnswizzleImage(Tegra::MemoryManager& gpu_memory,
-                                            const ImageInfo& info, std::span<const u8> input,
+                                                                   GPUVAddr gpu_addr,
-                                            std::span<u8> output) {
+                                                                   const ImageInfo& info,
+                                                                   std::span<const u8> input,
+                                                                   std::span<u8> output) {
    const size_t guest_size_bytes = input.size_bytes();
    const u32 bpp_log2 = BytesPerBlockLog2(info.format);
    const Extent3D size = info.size;
@@ -861,7 +867,7 @@ std::vector<BufferImageCopy> UnswizzleImage(Tegra::MemoryManager& gpu_memory, GP
                                            info.tile_width_spacing);
    size_t guest_offset = 0;
    u32 host_offset = 0;
-    std::vector<BufferImageCopy> copies(num_levels);
+    boost::container::small_vector<BufferImageCopy, 16> copies(num_levels);
    for (s32 level = 0; level < num_levels; ++level) {
        const Extent3D level_size = AdjustMipSize(size, level);
@@ -978,7 +984,7 @@ void ConvertImage(std::span<const u8> input, const ImageInfo& info, std::span<u8
    }
 }
-std::vector<BufferImageCopy> FullDownloadCopies(const ImageInfo& info) {
+boost::container::small_vector<BufferImageCopy, 16> FullDownloadCopies(const ImageInfo& info) {
    const Extent3D size = info.size;
    const u32 bytes_per_block = BytesPerBlock(info.format);
    if (info.type == ImageType::Linear) {
@@ -1006,7 +1012,7 @@ std::vector<BufferImageCopy> FullDownloadCopies(const ImageInfo& info) {
    u32 host_offset = 0;
-    std::vector<BufferImageCopy> copies(num_levels);
+    boost::container::small_vector<BufferImageCopy, 16> copies(num_levels);
    for (s32 level = 0; level < num_levels; ++level) {
        const Extent3D level_size = AdjustMipSize(size, level);
        const u32 num_blocks_per_layer = NumBlocks(level_size, tile_size);
@@ -1042,10 +1048,10 @@ Extent3D MipBlockSize(const ImageInfo& info, u32 level) {
    return AdjustMipBlockSize(num_tiles, level_info.block, level);
 }
-std::vector<SwizzleParameters> FullUploadSwizzles(const ImageInfo& info) {
+boost::container::small_vector<SwizzleParameters, 16> FullUploadSwizzles(const ImageInfo& info) {
    const Extent2D tile_size = DefaultBlockSize(info.format);
    if (info.type == ImageType::Linear) {
-        return std::vector{SwizzleParameters{
+        return {SwizzleParameters{
            .num_tiles = AdjustTileSize(info.size, tile_size),
            .block = {},
            .buffer_offset = 0,
@@ -1057,7 +1063,7 @@ std::vector<SwizzleParameters> FullUploadSwizzles(const ImageInfo& info) {
    const s32 num_levels = info.resources.levels;
    u32 guest_offset = 0;
-    std::vector<SwizzleParameters> params(num_levels);
+    boost::container::small_vector<SwizzleParameters, 16> params(num_levels);
    for (s32 level = 0; level < num_levels; ++level) {
        const Extent3D level_size = AdjustMipSize(size, level);
        const Extent3D num_tiles = AdjustTileSize(level_size, tile_size);
diff --git a/src/video_core/texture_cache/util.h b/src/video_core/texture_cache/util.h
index 84aa6880d..ab45a43c4 100644
--- a/src/video_core/texture_cache/util.h
+++ b/src/video_core/texture_cache/util.h
@@ -5,6 +5,7 @@
 #include <optional>
 #include <span>
+#include <boost/container/small_vector.hpp>
 #include "common/common_types.h"
 #include "common/scratch_buffer.h"
@@ -40,9 +41,10 @@ struct OverlapResult {
 [[nodiscard]] LevelArray CalculateMipLevelSizes(const ImageInfo& info) noexcept;
-[[nodiscard]] std::vector<u32> CalculateSliceOffsets(const ImageInfo& info);
+[[nodiscard]] boost::container::small_vector<u32, 16> CalculateSliceOffsets(const ImageInfo& info);
-[[nodiscard]] std::vector<SubresourceBase> CalculateSliceSubresources(const ImageInfo& info);
+[[nodiscard]] boost::container::small_vector<SubresourceBase, 16> CalculateSliceSubresources(
+    const ImageInfo& info);
 [[nodiscard]] u32 CalculateLevelStrideAlignment(const ImageInfo& info, u32 level);
@@ -51,21 +53,18 @@ struct OverlapResult {
 [[nodiscard]] ImageViewType RenderTargetImageViewType(const ImageInfo& info) noexcept;
-[[nodiscard]] std::vector<ImageCopy> MakeShrinkImageCopies(const ImageInfo& dst,
+[[nodiscard]] boost::container::small_vector<ImageCopy, 16> MakeShrinkImageCopies(
-                                                           const ImageInfo& src,
+    const ImageInfo& dst, const ImageInfo& src, SubresourceBase base, u32 up_scale = 1,
-                                                           SubresourceBase base, u32 up_scale = 1,
+    u32 down_shift = 0);
-                                                           u32 down_shift = 0);
-[[nodiscard]] std::vector<ImageCopy> MakeReinterpretImageCopies(const ImageInfo& src,
+[[nodiscard]] boost::container::small_vector<ImageCopy, 16> MakeReinterpretImageCopies(
-                                                                u32 up_scale = 1,
+    const ImageInfo& src, u32 up_scale = 1, u32 down_shift = 0);
-                                                                u32 down_shift = 0);
 [[nodiscard]] bool IsValidEntry(const Tegra::MemoryManager& gpu_memory, const TICEntry& config);
-[[nodiscard]] std::vector<BufferImageCopy> UnswizzleImage(Tegra::MemoryManager& gpu_memory,
+[[nodiscard]] boost::container::small_vector<BufferImageCopy, 16> UnswizzleImage(
-                                                          GPUVAddr gpu_addr, const ImageInfo& info,
+    Tegra::MemoryManager& gpu_memory, GPUVAddr gpu_addr, const ImageInfo& info,
-                                                          std::span<const u8> input,
+    std::span<const u8> input, std::span<u8> output);
-                                                          std::span<u8> output);
 [[nodiscard]] BufferCopy UploadBufferCopy(Tegra::MemoryManager& gpu_memory, GPUVAddr gpu_addr,
                                          const ImageBase& image, std::span<u8> output);
@@ -73,13 +72,15 @@ struct OverlapResult {
 void ConvertImage(std::span<const u8> input, const ImageInfo& info, std::span<u8> output,
                  std::span<BufferImageCopy> copies);
-[[nodiscard]] std::vector<BufferImageCopy> FullDownloadCopies(const ImageInfo& info);
+[[nodiscard]] boost::container::small_vector<BufferImageCopy, 16> FullDownloadCopies(
+    const ImageInfo& info);
 [[nodiscard]] Extent3D MipSize(Extent3D size, u32 level);
 [[nodiscard]] Extent3D MipBlockSize(const ImageInfo& info, u32 level);
-[[nodiscard]] std::vector<SwizzleParameters> FullUploadSwizzles(const ImageInfo& info);
+[[nodiscard]] boost::container::small_vector<SwizzleParameters, 16> FullUploadSwizzles(
+    const ImageInfo& info);
 void SwizzleImage(Tegra::MemoryManager& gpu_memory, GPUVAddr gpu_addr, const ImageInfo& info,
                  std::span<const BufferImageCopy> copies, std::span<const u8> memory,
diff --git a/src/video_core/transform_feedback.cpp b/src/video_core/transform_feedback.cpp
index 155599316..1f353d2df 100644
--- a/src/video_core/transform_feedback.cpp
+++ b/src/video_core/transform_feedback.cpp
@@ -13,7 +13,7 @@
 namespace VideoCommon {
-std::vector<Shader::TransformFeedbackVarying> MakeTransformFeedbackVaryings(
+std::pair<std::array<Shader::TransformFeedbackVarying, 256>, u32> MakeTransformFeedbackVaryings(
    const TransformFeedbackState& state) {
    static constexpr std::array VECTORS{
        28U,  // gl_Position
@@ -62,7 +62,8 @@ std::vector<Shader::TransformFeedbackVarying> MakeTransformFeedbackVaryings(
        216U, // gl_TexCoord[6]
        220U, // gl_TexCoord[7]
    };
-    std::vector<Shader::TransformFeedbackVarying> xfb(256);
+    std::array<Shader::TransformFeedbackVarying, 256> xfb{};
+    u32 count{0};
    for (size_t buffer = 0; buffer < state.layouts.size(); ++buffer) {
        const auto& locations = state.varyings[buffer];
        const auto& layout = state.layouts[buffer];
@@ -103,11 +104,12 @@ std::vector<Shader::TransformFeedbackVarying> MakeTransformFeedbackVaryings(
                }
            }
            xfb[attribute] = varying;
+            count = std::max(count, attribute);
            highest = std::max(highest, (base_offset + varying.components) * 4);
        }
        UNIMPLEMENTED_IF(highest != layout.stride);
    }
-    return xfb;
+    return {xfb, count + 1};
 }
 } // namespace VideoCommon
diff --git a/src/video_core/transform_feedback.h b/src/video_core/transform_feedback.h
index d13eb16c3..401b1352a 100644
--- a/src/video_core/transform_feedback.h
+++ b/src/video_core/transform_feedback.h
@@ -24,7 +24,7 @@ struct TransformFeedbackState {
        varyings;
 };
-std::vector<Shader::TransformFeedbackVarying> MakeTransformFeedbackVaryings(
+std::pair<std::array<Shader::TransformFeedbackVarying, 256>, u32> MakeTransformFeedbackVaryings(
    const TransformFeedbackState& state);
 } // namespace VideoCommon
diff --git a/src/video_core/vulkan_common/vulkan_device.cpp b/src/video_core/vulkan_common/vulkan_device.cpp
index fa9cde75b..b11abe311 100644
--- a/src/video_core/vulkan_common/vulkan_device.cpp
+++ b/src/video_core/vulkan_common/vulkan_device.cpp
@@ -316,6 +316,7 @@ NvidiaArchitecture GetNvidiaArchitecture(vk::PhysicalDevice physical,
 std::vector<const char*> ExtensionListForVulkan(
    const std::set<std::string, std::less<>>& extensions) {
    std::vector<const char*> output;
+    output.reserve(extensions.size());
    for (const auto& extension : extensions) {
        output.push_back(extension.c_str());
    }
diff --git a/src/yuzu/bootmanager.cpp b/src/yuzu/bootmanager.cpp
index cc6b6a25a..bdd1497b5 100644
--- a/src/yuzu/bootmanager.cpp
+++ b/src/yuzu/bootmanager.cpp
@@ -105,14 +105,12 @@ void EmuThread::run() {
        std::unique_lock lk{m_should_run_mutex};
        if (m_should_run) {
            m_system.Run();
-            m_is_running.store(true);
+            m_stopped.Reset();
-            m_is_running.notify_all();
            Common::CondvarWait(m_should_run_cv, lk, stop_token, [&] { return !m_should_run; });
        } else {
            m_system.Pause();
-            m_is_running.store(false);
+            m_stopped.Set();
-            m_is_running.notify_all();
            EmulationPaused(lk);
            Common::CondvarWait(m_should_run_cv, lk, stop_token, [&] { return m_should_run; });
diff --git a/src/yuzu/bootmanager.h b/src/yuzu/bootmanager.h
index b7b9d4141..87b23df12 100644
--- a/src/yuzu/bootmanager.h
+++ b/src/yuzu/bootmanager.h
@@ -3,7 +3,6 @@
 #pragma once
-#include <atomic>
 #include <condition_variable>
 #include <cstddef>
 #include <memory>
@@ -88,7 +87,7 @@ public:
        // Wait until paused, if pausing.
        if (!should_run) {
-            m_is_running.wait(true);
+            m_stopped.Wait();
        }
    }
@@ -97,7 +96,7 @@ public:
     * @return True if the emulation thread is running, otherwise false
     */
    bool IsRunning() const {
-        return m_is_running.load() || m_should_run;
+        return m_should_run;
    }
    /**
@@ -118,7 +117,7 @@ private:
    std::stop_source m_stop_source;
    std::mutex m_should_run_mutex;
    std::condition_variable_any m_should_run_cv;
-    std::atomic<bool> m_is_running{false};
+    Common::Event m_stopped;
    bool m_should_run{true};
 signals:
diff --git a/src/yuzu/configuration/config.cpp b/src/yuzu/configuration/config.cpp
index bac9dff90..edc206a25 100644
--- a/src/yuzu/configuration/config.cpp
+++ b/src/yuzu/configuration/config.cpp
@@ -761,6 +761,7 @@ void Config::ReadRendererValues() {
    ReadGlobalSetting(Settings::values.use_vulkan_driver_pipeline_cache);
    ReadGlobalSetting(Settings::values.enable_compute_pipelines);
    ReadGlobalSetting(Settings::values.use_video_framerate);
+    ReadGlobalSetting(Settings::values.barrier_feedback_loops);
    ReadGlobalSetting(Settings::values.bg_red);
    ReadGlobalSetting(Settings::values.bg_green);
    ReadGlobalSetting(Settings::values.bg_blue);
@@ -1417,6 +1418,7 @@ void Config::SaveRendererValues() {
    WriteGlobalSetting(Settings::values.use_vulkan_driver_pipeline_cache);
    WriteGlobalSetting(Settings::values.enable_compute_pipelines);
    WriteGlobalSetting(Settings::values.use_video_framerate);
+    WriteGlobalSetting(Settings::values.barrier_feedback_loops);
    WriteGlobalSetting(Settings::values.bg_red);
    WriteGlobalSetting(Settings::values.bg_green);
    WriteGlobalSetting(Settings::values.bg_blue);
diff --git a/src/yuzu/configuration/configure_graphics_advanced.cpp b/src/yuzu/configuration/configure_graphics_advanced.cpp
index 0463ac8b9..c0a044767 100644
--- a/src/yuzu/configuration/configure_graphics_advanced.cpp
+++ b/src/yuzu/configuration/configure_graphics_advanced.cpp
@@ -43,6 +43,8 @@ void ConfigureGraphicsAdvanced::SetConfiguration() {
    ui->enable_compute_pipelines_checkbox->setChecked(
        Settings::values.enable_compute_pipelines.GetValue());
    ui->use_video_framerate_checkbox->setChecked(Settings::values.use_video_framerate.GetValue());
+    ui->barrier_feedback_loops_checkbox->setChecked(
+        Settings::values.barrier_feedback_loops.GetValue());
    if (Settings::IsConfiguringGlobal()) {
        ui->gpu_accuracy->setCurrentIndex(
@@ -94,6 +96,9 @@ void ConfigureGraphicsAdvanced::ApplyConfiguration() {
                                             enable_compute_pipelines);
    ConfigurationShared::ApplyPerGameSetting(&Settings::values.use_video_framerate,
                                             ui->use_video_framerate_checkbox, use_video_framerate);
+    ConfigurationShared::ApplyPerGameSetting(&Settings::values.barrier_feedback_loops,
+                                             ui->barrier_feedback_loops_checkbox,
+                                             barrier_feedback_loops);
 }
 void ConfigureGraphicsAdvanced::changeEvent(QEvent* event) {
@@ -130,6 +135,8 @@ void ConfigureGraphicsAdvanced::SetupPerGameUI() {
            Settings::values.enable_compute_pipelines.UsingGlobal());
        ui->use_video_framerate_checkbox->setEnabled(
            Settings::values.use_video_framerate.UsingGlobal());
+        ui->barrier_feedback_loops_checkbox->setEnabled(
+            Settings::values.barrier_feedback_loops.UsingGlobal());
        return;
    }
@@ -157,6 +164,9 @@ void ConfigureGraphicsAdvanced::SetupPerGameUI() {
    ConfigurationShared::SetColoredTristate(ui->use_video_framerate_checkbox,
                                            Settings::values.use_video_framerate,
                                            use_video_framerate);
+    ConfigurationShared::SetColoredTristate(ui->barrier_feedback_loops_checkbox,
+                                            Settings::values.barrier_feedback_loops,
+                                            barrier_feedback_loops);
    ConfigurationShared::SetColoredComboBox(
        ui->gpu_accuracy, ui->label_gpu_accuracy,
        static_cast<int>(Settings::values.gpu_accuracy.GetValue(true)));
diff --git a/src/yuzu/configuration/configure_graphics_advanced.h b/src/yuzu/configuration/configure_graphics_advanced.h
index a4dc8ceb0..369a7c83e 100644
--- a/src/yuzu/configuration/configure_graphics_advanced.h
+++ b/src/yuzu/configuration/configure_graphics_advanced.h
@@ -48,6 +48,7 @@ private:
    ConfigurationShared::CheckState use_vulkan_driver_pipeline_cache;
    ConfigurationShared::CheckState enable_compute_pipelines;
    ConfigurationShared::CheckState use_video_framerate;
+    ConfigurationShared::CheckState barrier_feedback_loops;
    const Core::System& system;
 };
diff --git a/src/yuzu/configuration/configure_graphics_advanced.ui b/src/yuzu/configuration/configure_graphics_advanced.ui
index e7f0ef6be..d527a6f38 100644
--- a/src/yuzu/configuration/configure_graphics_advanced.ui
+++ b/src/yuzu/configuration/configure_graphics_advanced.ui
@@ -202,6 +202,16 @@ Compute pipelines are always enabled on all other drivers.</string>
         </widget>
        </item>
        <item>
+         <widget class="QCheckBox" name="barrier_feedback_loops_checkbox">
+          <property name="toolTip">
+           <string>Improves rendering of transparency effects in specific games.</string>
+          </property>
+          <property name="text">
+           <string>Barrier feedback loops</string>
+          </property>
+         </widget>
+        </item>
+        <item>
         <widget class="QWidget" name="af_layout" native="true">
          <layout class="QHBoxLayout" name="horizontalLayout_1">
           <property name="leftMargin">
diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp
index 8768a7c3c..45a39451d 100644
--- a/src/yuzu/main.cpp
+++ b/src/yuzu/main.cpp
@@ -3840,7 +3840,7 @@ void GMainWindow::OnLoadAmiibo() {
    auto* virtual_amiibo = input_subsystem->GetVirtualAmiibo();
    // Remove amiibo if one is connected
-    if (virtual_amiibo->GetCurrentState() == InputCommon::VirtualAmiibo::State::AmiiboIsOpen) {
+    if (virtual_amiibo->GetCurrentState() == InputCommon::VirtualAmiibo::State::TagNearby) {
        virtual_amiibo->CloseAmiibo();
        QMessageBox::warning(this, tr("Amiibo"), tr("The current amiibo has been removed"));
        return;
@@ -3868,7 +3868,7 @@ void GMainWindow::LoadAmiibo(const QString& filename) {
    auto* virtual_amiibo = input_subsystem->GetVirtualAmiibo();
    const QString title = tr("Error loading Amiibo data");
    // Remove amiibo if one is connected
-    if (virtual_amiibo->GetCurrentState() == InputCommon::VirtualAmiibo::State::AmiiboIsOpen) {
+    if (virtual_amiibo->GetCurrentState() == InputCommon::VirtualAmiibo::State::TagNearby) {
        virtual_amiibo->CloseAmiibo();
        QMessageBox::warning(this, tr("Amiibo"), tr("The current amiibo has been removed"));
        return;