VFCMULCPH/VFMULCPH—Complex Multiply FP16 Values

Opcode/Instruction Op/En 64/32 bit Mode Support CPUID Feature Flag Description

EVEX.128.F2.MAP6.W0 D6 /r

VFCMULCPH xmm1{k1}{z}, xmm2, xmm3/m128/m32bcst

A V/V AVX512-FP16 AVX512VL Complex multiply a pair of FP16 values from xmm2 and xmm3/m128/m32bcst, and store the result in xmm1 subject to writemask k1.

EVEX.256.F2.MAP6.W0 D6 /r

VFCMULCPH ymm1{k1}{z}, ymm2, ymm3/m256/m32bcst

A V/V AVX512-FP16 AVX512VL Complex multiply a pair of FP16 values from ymm2 and ymm3/m256/m32bcst, and store the result in ymm1 subject to writemask k1.
EVEX.512.F2.MAP6.W0 D6 /r VFCMULCPH zmm1{k1}{z}, zmm2, zmm3/m512/m32bcst {er} A V/V AVX512-FP16 Complex multiply a pair of FP16 values from zmm2 and zmm3/m512/m32bcst, and store the result in zmm1 subject to writemask k1.

EVEX.128.F3.MAP6.W0 D6 /r

VFMULCPH xmm1{k1}{z}, xmm2, xmm3/m128/m32bcst

A V/V AVX512-FP16 AVX512VL Complex multiply a pair of FP16 values from xmm2 and the complex conjugate of xmm3/m128/m32bcst, and store the result in xmm1 subject to writemask k1.

EVEX.256.F3.MAP6.W0 D6 /r

VFMULCPH ymm1{k1}{z}, ymm2, ymm3/m256/m32bcst

A V/V AVX512-FP16 AVX512VL Complex multiply a pair of FP16 values from ymm2 and the complex conjugate of ymm3/m256/m32bcst, and store the result in ymm1 subject to writemask k1.
EVEX.512.F3.MAP6.W0 D6 /r VFMULCPH zmm1{k1}{z}, zmm2, zmm3/m512/m32bcst {er} A V/V AVX512-FP16 Complex multiply a pair of FP16 values from zmm2 and the complex conjugate of zmm3/m512/m32bcst, and store the result in zmm1 subject to writemask k1.

Instruction Operand Encoding

Op/En Tuple Operand 1 Operand 2 Operand 3 Operand 4
A Full ModRM:reg (w) VEX.vvvv (r) ModRM:r/m (r) N/A

Description

This instruction performs a complex multiply operation. There are normal and complex conjugate forms of the oper-ation. The broadcasting and masking for this operation is done on 32-bit quantities representing a pair of FP16 values.

Rounding is performed at every FMA (fused multiply and add) boundary. Execution occurs as if all MXCSR excep-tions are masked. MXCSR status bits are updated to reflect exceptional conditions.

Operation

VFMULCPH dest{k1}, src1, src2 (AVX512)

VL = 128, 256 or 512

KL := VL/32

FOR i := 0 to KL-1:

IF k1[i] or *no writemask*:

IF broadcasting and src2 is memory:

tsrc2.fp16[2*i+0] := src2.fp16[0]

tsrc2.fp16[2*i+1] := src2.fp16[1]

ELSE:

tsrc2.fp16[2*i+0] := src2.fp16[2*i+0]

tsrc2.fp16[2*i+1] := src2.fp16[2*i+1]

FOR i := 0 to kl-1:

IF k1[i] or *no writemask*:

tmp.fp16[2*i+0] := src1.fp16[2*i+0] * tsrc2.fp16[2*i+0]

tmp.fp16[2*i+1] := src1.fp16[2*i+1] * tsrc2.fp16[2*i+0]

FOR i := 0 to KL-1:

IF k1[i] or *no writemask*:

// non-conjugate version subtracts last even term

dest.fp16[2*i+0] := tmp.fp16[2*i+0] - src1.fp16[2*i+1] * tsrc2.fp16[2*i+1]

dest.fp16[2*i+1] := tmp.fp16[2*i+1] + src1.fp16[2*i+0] * tsrc2.fp16[2*i+1]

ELSE IF *zeroing*:

dest.fp16[2*i+0] := 0

dest.fp16[2*i+1] := 0

DEST[MAXVL-1:VL] := 0

VFCMULCPH dest{k1}, src1, src2 (AVX512)

VL = 128, 256 or 512

KL := VL/32

FOR i := 0 to KL-1:

IF k1[i] or *no writemask*:

IF broadcasting and src2 is memory:

tsrc2.fp16[2*i+0] := src2.fp16[0]

tsrc2.fp16[2*i+1] := src2.fp16[1]

ELSE:

tsrc2.fp16[2*i+0] := src2.fp16[2*i+0]

tsrc2.fp16[2*i+1] := src2.fp16[2*i+1]

FOR i := 0 to KL-1:

IF k1[i] or *no writemask*:

tmp.fp16[2*i+0] := src1.fp16[2*i+0] * tsrc2.fp16[2*i+0]

tmp.fp16[2*i+1] := src1.fp16[2*i+1] * tsrc2.fp16[2*i+0]

FOR i := 0 to KL-1:

IF k1[i] or *no writemask*:

// conjugate version subtracts odd final term

dest.fp16[2*i] := tmp.fp16[2*i+0] +src1.fp16[2*i+1] * tsrc2.fp16[2*i+1]

dest.fp16[2*i+1] := tmp.fp16[2*i+1] - src1.fp16[2*i+0] * tsrc2.fp16[2*i+1]

ELSE IF *zeroing*:

dest.fp16[2*i+0] := 0

dest.fp16[2*i+1] := 0

DEST[MAXVL-1:VL] := 0

Intel C/C++ Compiler Intrinsic Equivalent

VFCMULCPH __m128h _mm_cmul_pch (__m128h a, __m128h b);

VFCMULCPH __m128h _mm_mask_cmul_pch (__m128h src, __mmask8 k, __m128h a, __m128h b);

VFCMULCPH __m128h _mm_maskz_cmul_pch (__mmask8 k, __m128h a, __m128h b);

VFCMULCPH __m256h _mm256_cmul_pch (__m256h a, __m256h b);

VFCMULCPH __m256h _mm256_mask_cmul_pch (__m256h src, __mmask8 k, __m256h a, __m256h b);

VFCMULCPH __m256h _mm256_maskz_cmul_pch (__mmask8 k, __m256h a, __m256h b);

VFCMULCPH __m512h _mm512_cmul_pch (__m512h a, __m512h b);

VFCMULCPH __m512h _mm512_mask_cmul_pch (__m512h src, __mmask16 k, __m512h a, __m512h b);

VFCMULCPH __m512h _mm512_maskz_cmul_pch (__mmask16 k, __m512h a, __m512h b);

VFCMULCPH __m512h _mm512_cmul_round_pch (__m512h a, __m512h b, const int rounding);

VFCMULCPH __m512h _mm512_mask_cmul_round_pch (__m512h src, __mmask16 k, __m512h a, __m512h b, const int rounding);

VFCMULCPH __m512h _mm512_maskz_cmul_round_pch (__mmask16 k, __m512h a, __m512h b, const int rounding);

VFCMULCPH __m128h _mm_fcmul_pch (__m128h a, __m128h b);

VFCMULCPH __m128h _mm_mask_fcmul_pch (__m128h src, __mmask8 k, __m128h a, __m128h b);

VFCMULCPH __m128h _mm_maskz_fcmul_pch (__mmask8 k, __m128h a, __m128h b);

VFCMULCPH __m256h _mm256_fcmul_pch (__m256h a, __m256h b);

VFCMULCPH __m256h _mm256_mask_fcmul_pch (__m256h src, __mmask8 k, __m256h a, __m256h b);

VFCMULCPH __m256h _mm256_maskz_fcmul_pch (__mmask8 k, __m256h a, __m256h b);

VFCMULCPH __m512h _mm512_fcmul_pch (__m512h a, __m512h b);

VFCMULCPH __m512h _mm512_mask_fcmul_pch (__m512h src, __mmask16 k, __m512h a, __m512h b);

VFCMULCPH __m512h _mm512_maskz_fcmul_pch (__mmask16 k, __m512h a, __m512h b);

VFCMULCPH __m512h _mm512_fcmul_round_pch (__m512h a, __m512h b, const int rounding);

VFCMULCPH __m512h _mm512_mask_fcmul_round_pch (__m512h src, __mmask16 k, __m512h a, __m512h b, const int rounding);

VFCMULCPH __m512h _mm512_maskz_fcmul_round_pch (__mmask16 k, __m512h a, __m512h b, const int rounding);

VFMULCPH __m128h _mm_fmul_pch (__m128h a, __m128h b);

VFMULCPH __m128h _mm_mask_fmul_pch (__m128h src, __mmask8 k, __m128h a, __m128h b);

VFMULCPH __m128h _mm_maskz_fmul_pch (__mmask8 k, __m128h a, __m128h b);

VFMULCPH __m256h _mm256_fmul_pch (__m256h a, __m256h b);

VFMULCPH __m256h _mm256_mask_fmul_pch (__m256h src, __mmask8 k, __m256h a, __m256h b);

VFMULCPH __m256h _mm256_maskz_fmul_pch (__mmask8 k, __m256h a, __m256h b);

VFMULCPH __m512h _mm512_fmul_pch (__m512h a, __m512h b);

VFMULCPH __m512h _mm512_mask_fmul_pch (__m512h src, __mmask16 k, __m512h a, __m512h b);

VFMULCPH __m512h _mm512_maskz_fmul_pch (__mmask16 k, __m512h a, __m512h b);

VFMULCPH __m512h _mm512_fmul_round_pch (__m512h a, __m512h b, const int rounding);

VFMULCPH __m512h _mm512_mask_fmul_round_pch (__m512h src, __mmask16 k, __m512h a, __m512h b, const int rounding);

VFMULCPH __m512h _mm512_maskz_fmul_round_pch (__mmask16 k, __m512h a, __m512h b, const int rounding);

VFMULCPH __m128h _mm_mask_mul_pch (__m128h src, __mmask8 k, __m128h a, __m128h b);

VFMULCPH __m128h _mm_maskz_mul_pch (__mmask8 k, __m128h a, __m128h b);

VFMULCPH __m128h _mm_mul_pch (__m128h a, __m128h b);

VFMULCPH __m256h _mm256_mask_mul_pch (__m256h src, __mmask8 k, __m256h a, __m256h b);

VFMULCPH __m256h _mm256_maskz_mul_pch (__mmask8 k, __m256h a, __m256h b);

VFMULCPH __m256h _mm256_mul_pch (__m256h a, __m256h b);

VFMULCPH __m512h _mm512_mask_mul_pch (__m512h src, __mmask16 k, __m512h a, __m512h b);

VFMULCPH __m512h _mm512_maskz_mul_pch (__mmask16 k, __m512h a, __m512h b);

VFMULCPH __m512h _mm512_mul_pch (__m512h a, __m512h b);

VFMULCPH __m512h _mm512_mask_mul_round_pch (__m512h src, __mmask16 k, __m512h a, __m512h b, const int rounding);

VFMULCPH __m512h _mm512_maskz_mul_round_pch (__mmask16 k, __m512h a, __m512h b, const int rounding);

VFMULCPH __m512h _mm512_mul_round_pch (__m512h a, __m512h b, const int rounding);

SIMD Floating-Point Exceptions

Invalid, Underflow, Overflow, Precision, Denormal

Other Exceptions

EVEX-encoded instructions, see Table 2-49, “Type E4 Class Exception Conditions.”

Additionally:

#UD If (dest_reg == src1_reg) or (dest_reg == src2_reg).