diff --git a/src/common/CMakeLists.txt b/src/common/CMakeLists.txt
index d8b255a40..19dc223ed 100644
--- a/src/common/CMakeLists.txt
+++ b/src/common/CMakeLists.txt
@@ -19,6 +19,7 @@ add_library(common
   file_system.h
   image.cpp
   image.h
+  intrin.h
   hash_combine.h
   heap_array.h
   heterogeneous_containers.h
diff --git a/src/common/bitutils.h b/src/common/bitutils.h
index b7f96a395..3461e06e5 100644
--- a/src/common/bitutils.h
+++ b/src/common/bitutils.h
@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #pragma once
@@ -8,6 +8,130 @@
 #include <intrin.h>
 #endif
 
+// Zero-extending helper
+template<typename TReturn, typename TValue>
+ALWAYS_INLINE constexpr TReturn ZeroExtend(TValue value)
+{
+  return static_cast<TReturn>(static_cast<typename std::make_unsigned<TReturn>::type>(
+    static_cast<typename std::make_unsigned<TValue>::type>(value)));
+}
+// Sign-extending helper
+template<typename TReturn, typename TValue>
+ALWAYS_INLINE constexpr TReturn SignExtend(TValue value)
+{
+  return static_cast<TReturn>(
+    static_cast<typename std::make_signed<TReturn>::type>(static_cast<typename std::make_signed<TValue>::type>(value)));
+}
+
+// Type-specific helpers
+template<typename TValue>
+ALWAYS_INLINE constexpr u16 ZeroExtend16(TValue value)
+{
+  return ZeroExtend<u16, TValue>(value);
+}
+template<typename TValue>
+ALWAYS_INLINE constexpr u32 ZeroExtend32(TValue value)
+{
+  return ZeroExtend<u32, TValue>(value);
+}
+template<typename TValue>
+ALWAYS_INLINE constexpr u64 ZeroExtend64(TValue value)
+{
+  return ZeroExtend<u64, TValue>(value);
+}
+template<typename TValue>
+ALWAYS_INLINE constexpr u16 SignExtend16(TValue value)
+{
+  return SignExtend<u16, TValue>(value);
+}
+template<typename TValue>
+ALWAYS_INLINE constexpr u32 SignExtend32(TValue value)
+{
+  return SignExtend<u32, TValue>(value);
+}
+template<typename TValue>
+ALWAYS_INLINE constexpr u64 SignExtend64(TValue value)
+{
+  return SignExtend<u64, TValue>(value);
+}
+template<typename TValue>
+ALWAYS_INLINE constexpr u8 Truncate8(TValue value)
+{
+  return static_cast<u8>(static_cast<typename std::make_unsigned<decltype(value)>::type>(value));
+}
+template<typename TValue>
+ALWAYS_INLINE constexpr u16 Truncate16(TValue value)
+{
+  return static_cast<u16>(static_cast<typename std::make_unsigned<decltype(value)>::type>(value));
+}
+template<typename TValue>
+ALWAYS_INLINE constexpr u32 Truncate32(TValue value)
+{
+  return static_cast<u32>(static_cast<typename std::make_unsigned<decltype(value)>::type>(value));
+}
+
+// BCD helpers
+ALWAYS_INLINE constexpr u8 BinaryToBCD(u8 value)
+{
+  return ((value / 10) << 4) + (value % 10);
+}
+ALWAYS_INLINE constexpr u8 PackedBCDToBinary(u8 value)
+{
+  return ((value >> 4) * 10) + (value % 16);
+}
+ALWAYS_INLINE constexpr u8 IsValidBCDDigit(u8 digit)
+{
+  return (digit <= 9);
+}
+ALWAYS_INLINE constexpr u8 IsValidPackedBCD(u8 value)
+{
+  return IsValidBCDDigit(value & 0x0F) && IsValidBCDDigit(value >> 4);
+}
+
+// Boolean to integer
+ALWAYS_INLINE constexpr u8 BoolToUInt8(bool value)
+{
+  return static_cast<u8>(value);
+}
+ALWAYS_INLINE constexpr u16 BoolToUInt16(bool value)
+{
+  return static_cast<u16>(value);
+}
+ALWAYS_INLINE constexpr u32 BoolToUInt32(bool value)
+{
+  return static_cast<u32>(value);
+}
+ALWAYS_INLINE constexpr u64 BoolToUInt64(bool value)
+{
+  return static_cast<u64>(value);
+}
+
+// Integer to boolean
+template<typename TValue>
+ALWAYS_INLINE constexpr bool ConvertToBool(TValue value)
+{
+  return static_cast<bool>(value);
+}
+
+// Unsafe integer to boolean
+template<typename TValue>
+ALWAYS_INLINE bool ConvertToBoolUnchecked(TValue value)
+{
+  // static_assert(sizeof(uint8) == sizeof(bool));
+  bool ret;
+  std::memcpy(&ret, &value, sizeof(bool));
+  return ret;
+}
+
+// Generic sign extension
+template<int NBITS, typename T>
+ALWAYS_INLINE constexpr T SignExtendN(T value)
+{
+  // http://graphics.stanford.edu/~seander/bithacks.html#VariableSignExtend
+  constexpr int shift = 8 * sizeof(T) - NBITS;
+  return static_cast<T>((static_cast<std::make_signed_t<T>>(value) << shift) >> shift);
+}
+
 /// Returns the number of zero bits before the first set bit, going MSB->LSB.
 template<typename T>
 ALWAYS_INLINE unsigned CountLeadingZeros(T value)
diff --git a/src/common/common.vcxproj b/src/common/common.vcxproj
index 1d7fee153..31b954b27 100644
--- a/src/common/common.vcxproj
+++ b/src/common/common.vcxproj
@@ -20,6 +20,7 @@
     <ClInclude Include="http_downloader.h" />
     <ClInclude Include="http_downloader_winhttp.h" />
     <ClInclude Include="image.h" />
+    <ClInclude Include="intrin.h" />
     <ClInclude Include="layered_settings_interface.h" />
     <ClInclude Include="log.h" />
     <ClInclude Include="lru_cache.h" />
diff --git a/src/common/common.vcxproj.filters b/src/common/common.vcxproj.filters
index 166b733f3..7f7a16267 100644
--- a/src/common/common.vcxproj.filters
+++ b/src/common/common.vcxproj.filters
@@ -43,6 +43,7 @@
     <ClInclude Include="sha1_digest.h" />
     <ClInclude Include="fastjmp.h" />
     <ClInclude Include="memmap.h" />
+    <ClInclude Include="intrin.h" />
   </ItemGroup>
   <ItemGroup>
     <ClCompile Include="small_string.cpp" />
diff --git a/src/common/intrin.h b/src/common/intrin.h
new file mode 100644
index 000000000..61595d297
--- /dev/null
+++ b/src/common/intrin.h
@@ -0,0 +1,26 @@
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
+
+// Includes appropriate intrinsic header based on platform.
+
+#pragma once
+
+#include "types.h"
+
+#if defined(CPU_ARCH_X86) || defined(CPU_ARCH_X64)
+#define CPU_ARCH_SSE 1
+#include <emmintrin.h>
+#elif defined(CPU_ARCH_ARM64)
+#define CPU_ARCH_NEON 1
+#ifdef _MSC_VER
+#include <arm64_neon.h>
+#else
+#include <arm_neon.h>
+#endif
+#endif
+
+#ifdef __APPLE__
+#include <stdlib.h> // alloca
+#else
+#include <malloc.h> // alloca
+#endif
diff --git a/src/common/types.h b/src/common/types.h
index 9a12b0b6e..540a1b608 100644
--- a/src/common/types.h
+++ b/src/common/types.h
@@ -180,130 +180,6 @@ static constexpr u32 HOST_PAGE_MASK = HOST_PAGE_SIZE - 1;
 static constexpr u32 HOST_PAGE_SHIFT = 12;
 #endif
 
-// Zero-extending helper
-template<typename TReturn, typename TValue>
-ALWAYS_INLINE constexpr TReturn ZeroExtend(TValue value)
-{
-  return static_cast<TReturn>(static_cast<typename std::make_unsigned<TReturn>::type>(
-    static_cast<typename std::make_unsigned<TValue>::type>(value)));
-}
-// Sign-extending helper
-template<typename TReturn, typename TValue>
-ALWAYS_INLINE constexpr TReturn SignExtend(TValue value)
-{
-  return static_cast<TReturn>(
-    static_cast<typename std::make_signed<TReturn>::type>(static_cast<typename std::make_signed<TValue>::type>(value)));
-}
-
-// Type-specific helpers
-template<typename TValue>
-ALWAYS_INLINE constexpr u16 ZeroExtend16(TValue value)
-{
-  return ZeroExtend<u16, TValue>(value);
-}
-template<typename TValue>
-ALWAYS_INLINE constexpr u32 ZeroExtend32(TValue value)
-{
-  return ZeroExtend<u32, TValue>(value);
-}
-template<typename TValue>
-ALWAYS_INLINE constexpr u64 ZeroExtend64(TValue value)
-{
-  return ZeroExtend<u64, TValue>(value);
-}
-template<typename TValue>
-ALWAYS_INLINE constexpr u16 SignExtend16(TValue value)
-{
-  return SignExtend<u16, TValue>(value);
-}
-template<typename TValue>
-ALWAYS_INLINE constexpr u32 SignExtend32(TValue value)
-{
-  return SignExtend<u32, TValue>(value);
-}
-template<typename TValue>
-ALWAYS_INLINE constexpr u64 SignExtend64(TValue value)
-{
-  return SignExtend<u64, TValue>(value);
-}
-template<typename TValue>
-ALWAYS_INLINE constexpr u8 Truncate8(TValue value)
-{
-  return static_cast<u8>(static_cast<typename std::make_unsigned<decltype(value)>::type>(value));
-}
-template<typename TValue>
-ALWAYS_INLINE constexpr u16 Truncate16(TValue value)
-{
-  return static_cast<u16>(static_cast<typename std::make_unsigned<decltype(value)>::type>(value));
-}
-template<typename TValue>
-ALWAYS_INLINE constexpr u32 Truncate32(TValue value)
-{
-  return static_cast<u32>(static_cast<typename std::make_unsigned<decltype(value)>::type>(value));
-}
-
-// BCD helpers
-ALWAYS_INLINE constexpr u8 BinaryToBCD(u8 value)
-{
-  return ((value / 10) << 4) + (value % 10);
-}
-ALWAYS_INLINE constexpr u8 PackedBCDToBinary(u8 value)
-{
-  return ((value >> 4) * 10) + (value % 16);
-}
-ALWAYS_INLINE constexpr u8 IsValidBCDDigit(u8 digit)
-{
-  return (digit <= 9);
-}
-ALWAYS_INLINE constexpr u8 IsValidPackedBCD(u8 value)
-{
-  return IsValidBCDDigit(value & 0x0F) && IsValidBCDDigit(value >> 4);
-}
-
-// Boolean to integer
-ALWAYS_INLINE constexpr u8 BoolToUInt8(bool value)
-{
-  return static_cast<u8>(value);
-}
-ALWAYS_INLINE constexpr u16 BoolToUInt16(bool value)
-{
-  return static_cast<u16>(value);
-}
-ALWAYS_INLINE constexpr u32 BoolToUInt32(bool value)
-{
-  return static_cast<u32>(value);
-}
-ALWAYS_INLINE constexpr u64 BoolToUInt64(bool value)
-{
-  return static_cast<u64>(value);
-}
-
-// Integer to boolean
-template<typename TValue>
-ALWAYS_INLINE constexpr bool ConvertToBool(TValue value)
-{
-  return static_cast<bool>(value);
-}
-
-// Unsafe integer to boolean
-template<typename TValue>
-ALWAYS_INLINE bool ConvertToBoolUnchecked(TValue value)
-{
-  // static_assert(sizeof(uint8) == sizeof(bool));
-  bool ret;
-  std::memcpy(&ret, &value, sizeof(bool));
-  return ret;
-}
-
-// Generic sign extension
-template<int NBITS, typename T>
-ALWAYS_INLINE constexpr T SignExtendN(T value)
-{
-  // http://graphics.stanford.edu/~seander/bithacks.html#VariableSignExtend
-  constexpr int shift = 8 * sizeof(T) - NBITS;
-  return static_cast<T>((static_cast<std::make_signed_t<T>>(value) << shift) >> shift);
-}
-
 // Enum class bitwise operators
 #define IMPLEMENT_ENUM_CLASS_BITWISE_OPERATORS(type_)                                                                  \
   ALWAYS_INLINE constexpr type_ operator&(type_ lhs, type_ rhs)                                                        \
diff --git a/src/core/analog_controller.cpp b/src/core/analog_controller.cpp
index 11389ba1c..9311b8d8e 100644
--- a/src/core/analog_controller.cpp
+++ b/src/core/analog_controller.cpp
@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com> and contributors.
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com> and contributors.
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "analog_controller.h"
@@ -10,6 +10,7 @@
 #include "util/input_manager.h"
 #include "util/state_wrapper.h"
 
+#include "common/bitutils.h"
 #include "common/log.h"
 #include "common/string_util.h"
 
diff --git a/src/core/analog_joystick.cpp b/src/core/analog_joystick.cpp
index 5bfc3bd35..b1ea88df3 100644
--- a/src/core/analog_joystick.cpp
+++ b/src/core/analog_joystick.cpp
@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com> and contributors.
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com> and contributors.
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "analog_joystick.h"
@@ -8,6 +8,7 @@
 #include "util/imgui_manager.h"
 #include "util/state_wrapper.h"
 
+#include "common/bitutils.h"
 #include "common/log.h"
 #include "common/string_util.h"
 
diff --git a/src/core/cdrom.cpp b/src/core/cdrom.cpp
index 78180fe59..ff20df621 100644
--- a/src/core/cdrom.cpp
+++ b/src/core/cdrom.cpp
@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "cdrom.h"
@@ -21,17 +21,15 @@
 #include "common/fifo_queue.h"
 #include "common/file_system.h"
 #include "common/heap_array.h"
+#include "common/intrin.h"
 #include "common/log.h"
 
 #include "imgui.h"
 
 #include <cmath>
 #include <vector>
-Log_SetChannel(CDROM);
 
-#if defined(CPU_ARCH_X64)
-#include <emmintrin.h>
-#endif
+Log_SetChannel(CDROM);
 
 namespace CDROM {
 enum : u32
@@ -3073,7 +3071,7 @@ static s16 GetPeakVolume(const u8* raw_sector, u8 channel)
 {
   static constexpr u32 NUM_SAMPLES = CDImage::RAW_SECTOR_SIZE / sizeof(s16);
 
-#if defined(CPU_ARCH_X64)
+#if defined(CPU_ARCH_SSE)
   static_assert(Common::IsAlignedPow2(NUM_SAMPLES, 8));
   const u8* current_ptr = raw_sector;
   __m128i v_peak = _mm_set1_epi16(0);
diff --git a/src/core/cpu_types.h b/src/core/cpu_types.h
index fb5297854..a4d0548a6 100644
--- a/src/core/cpu_types.h
+++ b/src/core/cpu_types.h
@@ -1,8 +1,9 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #pragma once
 #include "common/bitfield.h"
+#include "common/bitutils.h"
 #include "types.h"
 #include <optional>
 
diff --git a/src/core/digital_controller.cpp b/src/core/digital_controller.cpp
index a1657549c..02056ac58 100644
--- a/src/core/digital_controller.cpp
+++ b/src/core/digital_controller.cpp
@@ -1,12 +1,15 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com> and contributors.
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com> and contributors.
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "digital_controller.h"
-#include "common/assert.h"
 #include "host.h"
 #include "system.h"
+
 #include "util/state_wrapper.h"
 
+#include "common/assert.h"
+#include "common/bitutils.h"
+
 DigitalController::DigitalController(u32 index) : Controller(index)
 {
 }
diff --git a/src/core/gpu_sw.cpp b/src/core/gpu_sw.cpp
index d254a3a2e..1c37e3217 100644
--- a/src/core/gpu_sw.cpp
+++ b/src/core/gpu_sw.cpp
@@ -8,6 +8,7 @@
 
 #include "common/align.h"
 #include "common/assert.h"
+#include "common/intrin.h"
 #include "common/log.h"
 #include "common/make_array.h"
 
@@ -15,16 +16,6 @@
 
 Log_SetChannel(GPU_SW);
 
-#if defined(CPU_ARCH_X64)
-#include <emmintrin.h>
-#elif defined(CPU_ARCH_ARM64)
-#ifdef _MSC_VER
-#include <arm64_neon.h>
-#else
-#include <arm_neon.h>
-#endif
-#endif
-
 template<typename T>
 ALWAYS_INLINE static constexpr std::tuple<T, T> MinMax(T v1, T v2)
 {
@@ -162,7 +153,7 @@ ALWAYS_INLINE void CopyOutRow16<GPUTexture::Format::RGBA5551, u16>(const u16* sr
 {
   u32 col = 0;
 
-#if defined(CPU_ARCH_X64)
+#if defined(CPU_ARCH_SSE)
   const u32 aligned_width = Common::AlignDownPow2(width, 8);
   for (; col < aligned_width; col += 8)
   {
@@ -176,7 +167,7 @@ ALWAYS_INLINE void CopyOutRow16<GPUTexture::Format::RGBA5551, u16>(const u16* sr
     _mm_storeu_si128(reinterpret_cast<__m128i*>(dst_ptr), value);
     dst_ptr += 8;
   }
-#elif defined(CPU_ARCH_ARM64)
+#elif defined(CPU_ARCH_NEON)
   const u32 aligned_width = Common::AlignDownPow2(width, 8);
   for (; col < aligned_width; col += 8)
   {
@@ -201,7 +192,7 @@ ALWAYS_INLINE void CopyOutRow16<GPUTexture::Format::RGB565, u16>(const u16* src_
 {
   u32 col = 0;
 
-#if defined(CPU_ARCH_X64)
+#if defined(CPU_ARCH_SSE)
   const u32 aligned_width = Common::AlignDownPow2(width, 8);
   for (; col < aligned_width; col += 8)
   {
@@ -216,7 +207,7 @@ ALWAYS_INLINE void CopyOutRow16<GPUTexture::Format::RGB565, u16>(const u16* src_
     _mm_storeu_si128(reinterpret_cast<__m128i*>(dst_ptr), value);
     dst_ptr += 8;
   }
-#elif defined(CPU_ARCH_ARM64)
+#elif defined(CPU_ARCH_NEON)
   const u32 aligned_width = Common::AlignDownPow2(width, 8);
   const uint16x8_t single_mask = vdupq_n_u16(0x1F);
   for (; col < aligned_width; col += 8)
diff --git a/src/core/gpu_types.h b/src/core/gpu_types.h
index 75bba6e5d..e0cb11e31 100644
--- a/src/core/gpu_types.h
+++ b/src/core/gpu_types.h
@@ -1,8 +1,9 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #pragma once
 #include "common/bitfield.h"
+#include "common/bitutils.h"
 #include "common/rectangle.h"
 #include "types.h"
 #include <array>
diff --git a/src/core/imgui_overlays.cpp b/src/core/imgui_overlays.cpp
index 332ee420d..b1d8685ba 100644
--- a/src/core/imgui_overlays.cpp
+++ b/src/core/imgui_overlays.cpp
@@ -23,6 +23,7 @@
 #include "common/align.h"
 #include "common/assert.h"
 #include "common/file_system.h"
+#include "common/intrin.h"
 #include "common/log.h"
 #include "common/string_util.h"
 #include "common/timer.h"
@@ -41,16 +42,6 @@
 #include <span>
 #include <unordered_map>
 
-#if defined(CPU_ARCH_X64)
-#include <emmintrin.h>
-#elif defined(CPU_ARCH_ARM64)
-#ifdef _MSC_VER
-#include <arm64_neon.h>
-#else
-#include <arm_neon.h>
-#endif
-#endif
-
 Log_SetChannel(ImGuiManager);
 
 namespace ImGuiManager {
@@ -66,7 +57,7 @@ static void Draw();
 
 static std::tuple<float, float> GetMinMax(std::span<const float> values)
 {
-#if defined(CPU_ARCH_X64)
+#if defined(CPU_ARCH_SSE)
   __m128 vmin(_mm_loadu_ps(values.data()));
   __m128 vmax(vmin);
 
@@ -80,7 +71,7 @@ static std::tuple<float, float> GetMinMax(std::span<const float> values)
     vmax = _mm_max_ps(vmax, v);
   }
 
-#ifdef _MSC_VER
+#if defined(_MSC_VER) && !defined(__clang__)
   float min = std::min(vmin.m128_f32[0], std::min(vmin.m128_f32[1], std::min(vmin.m128_f32[2], vmin.m128_f32[3])));
   float max = std::max(vmax.m128_f32[0], std::max(vmax.m128_f32[1], std::max(vmax.m128_f32[2], vmax.m128_f32[3])));
 #else
@@ -94,7 +85,7 @@ static std::tuple<float, float> GetMinMax(std::span<const float> values)
   }
 
   return std::tie(min, max);
-#elif defined(CPU_ARCH_ARM64)
+#elif defined(CPU_ARCH_NEON)
   float32x4_t vmin(vld1q_f32(values.data()));
   float32x4_t vmax(vmin);
 
diff --git a/src/core/memory_card.cpp b/src/core/memory_card.cpp
index 4caf8f87f..8a757222a 100644
--- a/src/core/memory_card.cpp
+++ b/src/core/memory_card.cpp
@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "memory_card.h"
@@ -8,6 +8,7 @@
 #include "util/imgui_manager.h"
 #include "util/state_wrapper.h"
 
+#include "common/bitutils.h"
 #include "common/byte_stream.h"
 #include "common/file_system.h"
 #include "common/log.h"
diff --git a/src/core/memory_card_image.cpp b/src/core/memory_card_image.cpp
index 7d59a6414..f7bb26eb4 100644
--- a/src/core/memory_card_image.cpp
+++ b/src/core/memory_card_image.cpp
@@ -1,18 +1,23 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "memory_card_image.h"
+#include "system.h"
+
+#include "util/shiftjis.h"
+#include "util/state_wrapper.h"
+
+#include "common/bitutils.h"
 #include "common/byte_stream.h"
 #include "common/file_system.h"
 #include "common/log.h"
 #include "common/path.h"
 #include "common/string_util.h"
-#include "system.h"
-#include "util/shiftjis.h"
-#include "util/state_wrapper.h"
+
 #include <algorithm>
 #include <cstdio>
 #include <optional>
+
 Log_SetChannel(MemoryCard);
 
 namespace MemoryCardImage {
diff --git a/src/core/negcon.cpp b/src/core/negcon.cpp
index ae88f5103..df459c476 100644
--- a/src/core/negcon.cpp
+++ b/src/core/negcon.cpp
@@ -1,12 +1,16 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com> and contributors.
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com> and contributors.
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "negcon.h"
-#include "common/assert.h"
-#include "common/log.h"
 #include "host.h"
 #include "system.h"
+
 #include "util/state_wrapper.h"
+
+#include "common/assert.h"
+#include "common/bitutils.h"
+#include "common/log.h"
+
 #include <array>
 #include <cmath>
 
@@ -250,7 +254,7 @@ static const Controller::ControllerBindingInfo s_binding_info[] = {
   BUTTON("R", TRANSLATE_NOOP("NeGcon", "Right Trigger"), NeGcon::Button::R, GenericInputBinding::R1),
   AXIS("SteeringLeft", TRANSLATE_NOOP("NeGcon", "Steering (Twist) Left"), NeGcon::HalfAxis::SteeringLeft, GenericInputBinding::LeftStickLeft),
   AXIS("SteeringRight", TRANSLATE_NOOP("NeGcon", "Steering (Twist) Right"), NeGcon::HalfAxis::SteeringRight, GenericInputBinding::LeftStickRight),
-  // clang-format on
+// clang-format on
 
 #undef AXIS
 #undef BUTTON
@@ -265,12 +269,9 @@ static const SettingInfo s_settings[] = {
    nullptr, 100.0f},
 };
 
-const Controller::ControllerInfo NeGcon::INFO = {ControllerType::NeGcon,
-                                                 "NeGcon",
-                                                 TRANSLATE_NOOP("ControllerType", "NeGcon"),
-                                                 s_binding_info,
-                                                 s_settings,
-                                                 Controller::VibrationCapabilities::NoVibration};
+const Controller::ControllerInfo NeGcon::INFO = {
+  ControllerType::NeGcon, "NeGcon",   TRANSLATE_NOOP("ControllerType", "NeGcon"),
+  s_binding_info,         s_settings, Controller::VibrationCapabilities::NoVibration};
 
 void NeGcon::LoadSettings(SettingsInterface& si, const char* section)
 {
diff --git a/src/core/pad.cpp b/src/core/pad.cpp
index c302e84d3..3814f215b 100644
--- a/src/core/pad.cpp
+++ b/src/core/pad.cpp
@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com> and contributors.
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com> and contributors.
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "pad.h"
@@ -15,6 +15,7 @@
 #include "util/state_wrapper.h"
 
 #include "common/bitfield.h"
+#include "common/bitutils.h"
 #include "common/fifo_queue.h"
 #include "common/log.h"
 
diff --git a/src/core/sio.cpp b/src/core/sio.cpp
index 83de12c78..0b4b1952a 100644
--- a/src/core/sio.cpp
+++ b/src/core/sio.cpp
@@ -1,16 +1,21 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "sio.h"
-#include "common/bitfield.h"
-#include "common/fifo_queue.h"
-#include "common/log.h"
 #include "controller.h"
 #include "interrupt_controller.h"
 #include "memory_card.h"
+
 #include "util/state_wrapper.h"
+
+#include "common/bitfield.h"
+#include "common/bitutils.h"
+#include "common/fifo_queue.h"
+#include "common/log.h"
+
 #include <array>
 #include <memory>
+
 Log_SetChannel(SIO);
 
 namespace SIO {
diff --git a/src/core/spu.cpp b/src/core/spu.cpp
index 325eb3499..5b3167d98 100644
--- a/src/core/spu.cpp
+++ b/src/core/spu.cpp
@@ -1,4 +1,4 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "spu.h"
@@ -15,6 +15,7 @@
 #include "util/wav_writer.h"
 
 #include "common/bitfield.h"
+#include "common/bitutils.h"
 #include "common/fifo_queue.h"
 #include "common/log.h"
 #include "common/path.h"
diff --git a/src/core/texture_replacements.cpp b/src/core/texture_replacements.cpp
index d93b0aac2..c41469bdb 100644
--- a/src/core/texture_replacements.cpp
+++ b/src/core/texture_replacements.cpp
@@ -1,20 +1,25 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "texture_replacements.h"
+#include "host.h"
+#include "settings.h"
+
+#include "common/bitutils.h"
 #include "common/file_system.h"
 #include "common/log.h"
 #include "common/path.h"
 #include "common/string_util.h"
 #include "common/timer.h"
+
 #include "fmt/format.h"
-#include "host.h"
-#include "settings.h"
 #include "xxhash.h"
 #if defined(CPU_ARCH_X86) || defined(CPU_ARCH_X64)
 #include "xxh_x86dispatch.h"
 #endif
+
 #include <cinttypes>
+
 Log_SetChannel(TextureReplacements);
 
 TextureReplacements g_texture_replacements;
diff --git a/src/util/audio_stream.cpp b/src/util/audio_stream.cpp
index 797a4d698..0ccda38ac 100644
--- a/src/util/audio_stream.cpp
+++ b/src/util/audio_stream.cpp
@@ -2,30 +2,20 @@
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "audio_stream.h"
-#include "SoundTouch.h"
+
 #include "common/align.h"
 #include "common/assert.h"
+#include "common/intrin.h"
 #include "common/log.h"
 #include "common/make_array.h"
 #include "common/timer.h"
+
+#include "SoundTouch.h"
+
 #include <algorithm>
 #include <cmath>
 #include <cstring>
 
-#ifdef __APPLE__
-#include <stdlib.h> // alloca
-#else
-#include <malloc.h> // alloca
-#endif
-
-#if defined(_M_ARM64)
-#include <arm64_neon.h>
-#elif defined(__aarch64__)
-#include <arm_neon.h>
-#elif defined(CPU_ARCH_X86) || defined(CPU_ARCH_X64)
-#include <emmintrin.h>
-#endif
-
 Log_SetChannel(AudioStream);
 
 static constexpr bool LOG_TIMESTRETCH_STATS = false;
@@ -364,7 +354,7 @@ void AudioStream::EndWrite(u32 num_frames)
 static constexpr float S16_TO_FLOAT = 1.0f / 32767.0f;
 static constexpr float FLOAT_TO_S16 = 32767.0f;
 
-#if defined(CPU_ARCH_ARM64)
+#if defined(CPU_ARCH_NEON)
 
 static void S16ChunkToFloat(const s32* src, float* dst)
 {
@@ -417,7 +407,7 @@ static void FloatChunkToS16(s32* dst, const float* src, uint size)
   }
 }
 
-#elif defined(CPU_ARCH_X86) || defined(CPU_ARCH_X64)
+#elif defined(CPU_ARCH_SSE)
 
 static void S16ChunkToFloat(const s32* src, float* dst)
 {
diff --git a/src/util/cd_image.h b/src/util/cd_image.h
index fd4a5e6de..1ad169b7f 100644
--- a/src/util/cd_image.h
+++ b/src/util/cd_image.h
@@ -3,6 +3,7 @@
 
 #pragma once
 #include "common/bitfield.h"
+#include "common/bitutils.h"
 #include "common/progress_callback.h"
 #include "common/types.h"
 #include <array>
diff --git a/src/util/d3d11_device.cpp b/src/util/d3d11_device.cpp
index 85ee28981..147705006 100644
--- a/src/util/d3d11_device.cpp
+++ b/src/util/d3d11_device.cpp
@@ -9,6 +9,7 @@
 
 #include "common/align.h"
 #include "common/assert.h"
+#include "common/bitutils.h"
 #include "common/file_system.h"
 #include "common/log.h"
 #include "common/path.h"
diff --git a/src/util/gpu_texture.cpp b/src/util/gpu_texture.cpp
index a08ae9b5c..f5f635ea6 100644
--- a/src/util/gpu_texture.cpp
+++ b/src/util/gpu_texture.cpp
@@ -4,6 +4,7 @@
 #include "gpu_texture.h"
 #include "gpu_device.h"
 
+#include "common/bitutils.h"
 #include "common/log.h"
 #include "common/string_util.h"
 
diff --git a/src/util/shadergen.cpp b/src/util/shadergen.cpp
index c1b80b241..2c5508e10 100644
--- a/src/util/shadergen.cpp
+++ b/src/util/shadergen.cpp
@@ -1,8 +1,9 @@
-// SPDX-FileCopyrightText: 2019-2022 Connor McLaughlin <stenzek@gmail.com>
+// SPDX-FileCopyrightText: 2019-2023 Connor McLaughlin <stenzek@gmail.com>
 // SPDX-License-Identifier: (GPL-3.0 OR CC-BY-NC-ND-4.0)
 
 #include "shadergen.h"
 #include "common/assert.h"
+#include "common/bitutils.h"
 #include "common/log.h"
 #include <cstdio>
 #include <cstring>