ymm vectorization

//const __m128 half = _mm_set1_ps(0.5f);

//set our packed floats to 1.5f
__m256 half256ps = _mm256_set1_ps(1.5f);

//make room for our 256bits result variable
__m256i result256i;

unsigned X, X0, X1, X2, X3, X4, X5, X6, X7;

__asm {
	//movd xmm0, X
	
	//broadcast X (unsigned 32bit) into 256bit ymm0 register
	vpbroadcastd	ymm0, X
	//mov				ecx, 5				
	//vbroadcastss	ymm1, half
	
	//set ymm1 register to packed 1.5f
	vmovdqu			ymm1, half256ps
	//vaddpd			ymm2, ymm0, ymm0
	
	//simd multiply ymm2 = ymm0 * ymm1
	vmulps			ymm2, ymm0, ymm1
	//movd			X,  xmm2
	
	//store multiplication value into result256i
	vmovdqu			result256i, ymm2
}

//result256i is an union, this is how we print it.
printf("%u  %u  %u  %u  %u  %u  %u  %u\n", 
	result256i.m256i_u32[0], result256i.m256i_u32[1], 
	result256i.m256i_u32[2], result256i.m256i_u32[3], 
	result256i.m256i_u32[4], result256i.m256i_u32[5], 
	result256i.m256i_u32[6], result256i.m256i_u32[7] );