VCVTPH2PD - ConVerT Packed Half to Packed Double
FP16 -> double
VCVTPH2PD xmm1{k1}{z}, xmm2/m32/m16bcst (V5+FP16+VL
__m128d _mm_cvtph_pd(__m128h a)
__m128d _mm_mask_cvtph_pd(__m128d s, __mmask8 k, __m128h a)
__m128d _mm_maskz_cvtph_pd(__mmask8 k, __m128h a)
VCVTPH2PD ymm1{k1}{z}, xmm2/m64/m16bcst (V5+FP16+VL
__m256d _mm256_cvtph_pd(__m128h a)
__m256d _mm256_mask_cvtph_pd(__m256d s, __mmask8 k, __m128h a)
__m256d _mm256_maskz_cvtph_pd(__mmask8 k, __m128h a)
VCVTPH2PD zmm1{k1}{z}, xmm2/m128/m16bcst{sae} (V5+FP16
__m512d _mm512_cvtph_pd(__m128h a)
__m512d _mm512_mask_cvtph_pd(__m512d s, __mmask8 k, __m128h a)
__m512d _mm512_maskz_cvtph_pd(__mmask8 k, __m128h a)
__m512d _mm512_cvt_roundph_pd(__m128h a, int sae)
__m512d _mm512_mask_cvt_roundph_pd(__m512d s, __mmask8 k, __m128h a, int sae)
__m512d _mm512_maskz_cvt_roundph_pd(__mmask8 k, __m128h a, int sae)
x86/x64 SIMD Instruction List
Feedback