diff --git a/rtcore.h b/rtcore.h
index bdc4672..279393c 100644
--- a/rtcore.h
+++ b/rtcore.h
@@ -53,7 +53,15 @@
  * This assume x64, which is what I'm always using.
  * You still need to build with the appropriate compilation
  * flags (-march=avx2 etc. for gcc, /arch:AVX2 etc. for msvc) */
-#include <immintrin.h>
+#if defined(__has_include) && !defined(RTC_NO_SIMD)
+	#if !__has_include(<immintrin.h>)
+		#define RTC_NO_SIMD
+	#else
+		#include <immintrin.h>
+	#endif
+#elif !defined(RTC_NO_SIMD)
+	#include <immintrin.h>
+#endif
 /* MSVC additionally has intrin.h for microsoft specific intrinsics */
 #ifdef _MSC_VER
 	#include <intrin.h>
@@ -128,6 +136,9 @@ typedef int32_t b32;
 		}
 
   #define force_inline __forceinline
+#else
+	#define assert(x) 
+	#define force_inline inline
 #endif
 
 /* Arena allocator */
@@ -241,58 +252,66 @@ RTC_API s8 ReadEntireFileS8(s8 path, arena *a);
 RTC_API b32 WriteEntireFile(s8 path, byte *data, isize length);
 
 /* Atomics */
-#if defined(__GNUC__) || defined(__clang__)
-	/* Atomic add */
-	#define AtomicAdd32(_addend, _val) __atomic_add_fetch((i32 *)_addend, _val, __ATOMIC_SEQ_CST)
-	#define AtomicAdd64(_addend, _val) __atomic_add_fetch((i64 *)_addend, _val, __ATOMIC_SEQ_CST)
-	#define AtomicStore(_ptr, _val) __atomic_store_n(_ptr, _val, __ATOMIC_SEQ_CST)
-	#define AtomicStoreRelease(_ptr, _val) __atomic_store_n(_ptr, _val, __ATOMIC_RELEASE)
-	#define AtomicLoad(_ptr) __atomic_load_n(_ptr, __ATOMIC_SEQ_CST)
-	#define AtomicLoadAcquire(_ptr) __atomic_load_n(_ptr, __ATOMIC_ACQUIRE)
-#elif defined(_MSC_VER)
-	#define AtomicAdd32(_addend, _val) _InterlockedExchangeAdd((volatile long *)_addend, _val)
-	#define AtomicAdd64(_addend, _val) _InterlockedExchangeAdd64((volatile __int64 *)_addend, _val)
-	#define AtomicStore(_ptr, _val) _InterlockedExchange((volatile long *)_ptr, _val)
-	#define AtomicStoreRelease(_ptr, _val) _InterlockedExchange_HLERelease(_ptr, _val)
-	#define AtomicLoad(_ptr) _InterlockedOr(_ptr, 0)
-	#define AtomicLoadAcquire(_ptr) _InterlockedOr_HLEAcquire(_ptr, 0)
+#if !defined(RTC_NO_ATOMICS)
+	#if defined(__GNUC__) || defined(__clang__)
+		/* Atomic add */
+		#define AtomicAdd32(_addend, _val) __atomic_add_fetch((i32 *)_addend, _val, __ATOMIC_SEQ_CST)
+		#define AtomicAdd64(_addend, _val) __atomic_add_fetch((i64 *)_addend, _val, __ATOMIC_SEQ_CST)
+		#define AtomicStore(_ptr, _val) __atomic_store_n(_ptr, _val, __ATOMIC_SEQ_CST)
+		#define AtomicStoreRelease(_ptr, _val) __atomic_store_n(_ptr, _val, __ATOMIC_RELEASE)
+		#define AtomicLoad(_ptr) __atomic_load_n(_ptr, __ATOMIC_SEQ_CST)
+		#define AtomicLoadAcquire(_ptr) __atomic_load_n(_ptr, __ATOMIC_ACQUIRE)
+	#elif defined(_MSC_VER)
+		#define AtomicAdd32(_addend, _val) _InterlockedExchangeAdd((volatile long *)_addend, _val)
+		#define AtomicAdd64(_addend, _val) _InterlockedExchangeAdd64((volatile __int64 *)_addend, _val)
+		#define AtomicStore(_ptr, _val) _InterlockedExchange((volatile long *)_ptr, _val)
+		#define AtomicStoreRelease(_ptr, _val) _InterlockedExchange_HLERelease(_ptr, _val)
+		#define AtomicLoad(_ptr) _InterlockedOr(_ptr, 0)
+		#define AtomicLoadAcquire(_ptr) _InterlockedOr_HLEAcquire(_ptr, 0)
+	#else
+		#define RTC_NO_ATOMICS
+	#endif
 #endif
 
 /* Aliases for intrinsics */
-#if defined(__GNUC__) || defined(__clang__)
-	#define CTZ32(ui) __builtin_ctz(ui)
-	#define CTZ64(ul) __builtin_ctzl(ul)
-	#define CLZ32(ui) __builtin_clz(ui)
-	#define CLZ64(ul) __builtin_clzl(ul)
-	#define PopCount32(_x) __builtin_popcount(_x)
-	#define PopCount64(_x) __builtin_popcountl(_x)
-#elif defined(_MSC_VER)
-static force_inline unsigned int CTZ32(u32 x)
-{
-	unsigned int index;
-	_BitScanReverse(&index, x);
-	return index;
-}
-static force_inline unsigned int CTZ64(u64 x)
-{
-	unsigned int index;
-	_BitScanReverse64(&index, x);
-	return index;
-}
-static force_inline unsigned int CLZ32(u32 x)
-{
-	unsigned int index;
-	_BitScanForward(&index, x);
-	return index;
-}
-static force_inline unsigned int CLZ64(u64 x)
-{
-	unsigned int index;
-	_BitScanForward64(&index, x);
-	return index;
-}
-	#define PopCount32(_x) __popcnt(_x)
-	#define PopCount64(_x) __popcnt64(_x)
+#if !defined(RTC_NO_INTRINSICS)
+	#if defined(__GNUC__) || defined(__clang__)
+		#define CTZ32(ui) __builtin_ctz(ui)
+		#define CTZ64(ul) __builtin_ctzl(ul)
+		#define CLZ32(ui) __builtin_clz(ui)
+		#define CLZ64(ul) __builtin_clzl(ul)
+		#define PopCount32(_x) __builtin_popcount(_x)
+		#define PopCount64(_x) __builtin_popcountl(_x)
+	#elif defined(_MSC_VER)
+	static force_inline unsigned int CTZ32(u32 x)
+	{
+		unsigned int index;
+		_BitScanReverse(&index, x);
+		return index;
+	}
+	static force_inline unsigned int CTZ64(u64 x)
+	{
+		unsigned int index;
+		_BitScanReverse64(&index, x);
+		return index;
+	}
+	static force_inline unsigned int CLZ32(u32 x)
+	{
+		unsigned int index;
+		_BitScanForward(&index, x);
+		return index;
+	}
+	static force_inline unsigned int CLZ64(u64 x)
+	{
+		unsigned int index;
+		_BitScanForward64(&index, x);
+		return index;
+	}
+		#define PopCount32(_x) __popcnt(_x)
+		#define PopCount64(_x) __popcnt64(_x)
+	#else
+		#define RTC_NO_INTRINSICS
+	#endif
 #endif
 
 /* Threading wrapper */
diff --git a/test.c b/test.c
index 4d3ed96..5a08bbf 100644
--- a/test.c
+++ b/test.c
@@ -6,8 +6,13 @@
 THREAD_FN(TestThread)
 {
   int *p = param;
+#ifndef RTC_NO_ATOMICS
   printf("Thread got param %d\n", AtomicLoad(p));
   AtomicStore(p, 42);
+#else
+  printf("Thread go param %d\n", *p);
+  *p = 42;
+#endif
   return 0;
 }
 
@@ -44,7 +49,11 @@ ThreadTest(void)
   int p = 32;
   thread *t = StartThread(TestThread, &p);
   JoinThread(t);
+#ifndef RTC_NO_ATOMICS
   return AtomicLoad(&p) == 42;
+#else
+  return p == 42;
+#endif
 }
 
 internal int