VCVTW2PH - ConVerT Word to Packed Half
signed WORD -> FP16
VCVTW2PH xmm1{k1}{z}, xmm2/m128/m16bcst (V5+FP16+VL
__m128h _mm_cvtepi16_ph(__m128i a)
__m128h _mm_mask_cvtepi16_ph(__m128h s, __mmask8 k, __m128i a)
__m128h _mm_maskz_cvtepi16_ph(__mmask8 k, __m128i a)
VCVTW2PH ymm1{k1}{z}, ymm2/m256/m16bcst (V5+FP16+VL
__m256h _mm256_cvtepi16_ph(__m256i a)
__m256h _mm256_mask_cvtepi16_ph(__m256h s, __mmask16 k, __m256i a)
__m256h _mm256_maskz_cvtepi16_ph(__mmask16 k, __m256i a)
VCVTW2PH zmm1{k1}{z}, zmm2/m512/m16bcst{er} (V5+FP16
__m512h _mm512_cvtepi16_ph(__m512i a)
__m512h _mm512_mask_cvtepi16_ph(__m512h s, __mmask32 k, __m512i a)
__m512h _mm512_maskz_cvtepi16_ph(__mmask32 k, __m512i a)
__m512h _mm512_cvt_roundepi16_ph(__m512i a, int r)
__m512h _mm512_mask_cvt_roundepi16_ph(__m512h s, __mmask32 k, __m512i a, int r)
__m512h _mm512_maskz_cvt_roundepi16_ph(__mmask32 k, __m512i a, int r)
x86/x64 SIMD Instruction List
Feedback