Parent topic: Vector Intrinsics
- _mm512_alignr_epi32/ _mm512_mask_alignr_epi32
Shifts int32 elements right and concatenates vectors. Corresponding instruction isVALIGND. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_exp223_ps/ _mm512_mask_exp223_ps
Approximates the base-2 exponent of an int32 vector representing fixed point values with eight bits for sign and integer part, and 24 bits for the fraction. Corresponding instruction isVEXP223PS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_fixupnan_pd/ _mm512_mask_fixupnan_pd
Fixes up NaNs from float64 vectors. Corresponding instruction isVFIXUPNANPD. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_fixupnan_ps/ _mm512_mask_fixupnan_ps
Fixes up NaNs from float32 vectors. Corresponding instruction isVFIXUPNANPS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_getexp_ps/ _mm512_mask_getexp_ps
Extract float32 vector of exponents. Corresponding instruction isVGETEXPPS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_getexp_pd/ _mm512_mask_getexp_pd
Extract float64 vector of exponents. Corresponding instruction isVGETEXPPD. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_getmant_pd/ _mm512_mask_getmant_pd
Extract float64 normalized mantissas. Corresponding instruction isVGETMANTPD. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_getmant_ps/ _mm512_mask_getmant_ps
Extracts float32 normalized mantissas. Corresponding instruction isVGETMANTPS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_log2ae23_ps/ _mm512_mask_log2ae23_ps
Calculates base-2 logarithm of float32 vector, with absolute error bounded by 2^(-23). Corresponding instruction isVLOG2PS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_rcp23_ps/ _mm512_mask_rcp23_ps
Approximates the reciprocals of the float32 elements to 23 bits of accuracy. Corresponding instruction isVRCP23PS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_roundfxpnt_adjust_ps/ _mm512_mask_roundfxpnt_adjust_ps
Rounds float32 vector. Corresponding instruction isVRNDFXPNTPS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_roundfxpnt_adjust_pd/ _mm512_mask_roundfxpnt_adjust_pd
Rounds float64 vector. Corresponding instruction isVRNDFXPNTPD. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture). - _mm512_rsqrt23_ps/ _mm512_mask_rsqrt23_ps
Calculates reciprocal square root of float32 vector. Corresponding instruction isVRSQRT23PS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture).
Inglés