diff --git a/examples/ToyPathTracer/Source/MathSimd.h b/examples/ToyPathTracer/Source/MathSimd.h
index 3e0a5cbe..ef68ab9b 100644
--- a/examples/ToyPathTracer/Source/MathSimd.h
+++ b/examples/ToyPathTracer/Source/MathSimd.h
@@ -8,7 +8,7 @@
 
 #define kSimdWidth 4
 
-#if !defined(__arm__) && !defined(__arm64__) && !defined(__EMSCRIPTEN__)
+#if !defined(__arm__) && !defined(__arm64__) && !defined(__EMSCRIPTEN__) && !defined(_M_ARM64)
 
 // ---- SSE implementation
 
@@ -141,8 +141,14 @@ VM_INLINE float hmin(float4 v)
 // Returns a 4-bit code where bit0..bit3 is X..W
 VM_INLINE unsigned mask(float4 v)
 {
+#if defined(_M_ARM64)
+    static const uint32_t values[4] = { 1u, 2u, 4u, 8u };
+    const uint32x4_t movemask = vld1q_u32( values );
+    const uint32x4_t highbit = vdupq_n_u32( 0x80000000u );
+#else
     static const uint32x4_t movemask = { 1, 2, 4, 8 };
     static const uint32x4_t highbit = { 0x80000000, 0x80000000, 0x80000000, 0x80000000 };
+#endif
     uint32x4_t t0 = vreinterpretq_u32_f32(v.m);
     uint32x4_t t1 = vtstq_u32(t0, highbit);
     uint32x4_t t2 = vandq_u32(t1, movemask);
diff --git a/examples/ToyPathTracer/Source/Maths.h b/examples/ToyPathTracer/Source/Maths.h
index b587715f..790aabd0 100644
--- a/examples/ToyPathTracer/Source/Maths.h
+++ b/examples/ToyPathTracer/Source/Maths.h
@@ -12,7 +12,7 @@
 #if DO_FLOAT3_WITH_SIMD
 
 
-#if !defined(__arm__) && !defined(__arm64__)
+#if !defined(__arm__) && !defined(__arm64__) && !defined(_M_ARM64)
 
 // ---- SSE implementation
 
@@ -223,8 +223,14 @@ VM_INLINE float3 cross(float3 a, float3 b)
 // Returns a 3-bit code where bit0..bit2 is X..Z
 VM_INLINE unsigned mask(float3 v)
 {
+#if defined(_M_ARM64)
+    static const uint32_t values[4] = { 1u, 2u, 4u, 8u };
+    const uint32x4_t movemask = vld1q_u32( values );
+    const uint32x4_t highbit = vdupq_n_u32( 0x80000000u );
+#else 
     static const uint32x4_t movemask = { 1, 2, 4, 8 };
     static const uint32x4_t highbit = { 0x80000000, 0x80000000, 0x80000000, 0x80000000 };
+#endif
     uint32x4_t t0 = vreinterpretq_u32_f32(v.m);
     uint32x4_t t1 = vtstq_u32(t0, highbit);
     uint32x4_t t2 = vandq_u32(t1, movemask);
diff --git a/examples/ToyPathTracer/Source/enkiTS/TaskScheduler.cpp b/examples/ToyPathTracer/Source/enkiTS/TaskScheduler.cpp
index 187673ad..c3e86ed1 100644
--- a/examples/ToyPathTracer/Source/enkiTS/TaskScheduler.cpp
+++ b/examples/ToyPathTracer/Source/enkiTS/TaskScheduler.cpp
@@ -72,7 +72,11 @@ namespace
 		#if defined _M_IX86  || defined _M_X64
 			#pragma intrinsic(_mm_pause)
 			inline void Pause() { _mm_pause(); }
-		#endif
+        #elif defined(_M_ARM64)
+			inline void Pause() { __yield(); }
+        #else
+            inline void Pause() { /* No ops*/ }
+        #endif
 	#elif defined __i386__ || defined __x86_64__
 		inline void Pause() { __asm__ __volatile__("pause;"); }
 	#else
diff --git a/profiler/src/winmain.cpp b/profiler/src/winmain.cpp
index fd3c35f2..a2aa6958 100644
--- a/profiler/src/winmain.cpp
+++ b/profiler/src/winmain.cpp
@@ -14,6 +14,7 @@ int main( int argc, char** argv );
 
 int WINAPI WinMain( HINSTANCE hInstance, HINSTANCE hPrevInstance, LPSTR lpszCmd, int nCmd )
 {
+#if defined(_M_IX86) || defined(_M_X64)
     {
         uint32_t regs[4];
         __cpuidex( (int*)regs, 0, 0 );
@@ -42,7 +43,7 @@ int WINAPI WinMain( HINSTANCE hInstance, HINSTANCE hPrevInstance, LPSTR lpszCmd,
             return 0;
         }
     }
-
+#endif // x86/x64: AVX feature checks
     return main( __argc, __argv );
 }
 #endif
diff --git a/server/TracyPopcnt.hpp b/server/TracyPopcnt.hpp
index 07b98fd8..38b828a1 100644
--- a/server/TracyPopcnt.hpp
+++ b/server/TracyPopcnt.hpp
@@ -13,6 +13,24 @@ static inline uint64_t TracyLzcnt( uint64_t i )
 {
     return uint64_t( __builtin_clzll( i ) );
 }
+#elif defined(_M_ARM64)
+#include <arm_neon.h>
+static inline uint64_t TracyCountBits( uint64_t i )
+{
+    uint8x8_t v = vreinterpret_u8_u64( vdup_n_u64( i ) );
+    uint8x8_t popcnt = vcnt_u8( v );
+    return (uint64_t)vaddlv_u8( popcnt );
+}
+static inline uint64_t TracyLzcnt( uint64_t i )
+{
+    unsigned long index;
+    uint64_t leadingZeros = 64;
+    if( _BitScanReverse64( &index, i ) )
+    {
+        leadingZeros = leadingZeros - 1 - index;
+    }
+    return leadingZeros;
+}
 #elif defined _WIN64
 #  include <intrin.h>
 #  define TracyCountBits __popcnt64