vpshldq

Packed Shift Left Quadword Concatenate

VPSHLDQ zmm1 {k1}, zmm2, zmm3/m512, imm8

Funnel shift left of quadwords.

Details

Performs a funnel (concatenate and shift) left operation on pairs of 64-bit quadword elements: the result is (src1 || src2) shifted left by the count specified in imm8 modulo 64. This is an AVX-512-VBMI2 instruction operating on 8 quadword elements per 512-bit vector. Masking via k1 allows selective updates; no CPU arithmetic flags are modified.

Pseudocode Operation

for i ← 0 to 7 do { if (k1[i] or not_masked) { combined_hi ← src1.qword[i]; combined_lo ← src2.qword[i]; shift_count ← imm8 & 0x3F; if (shift_count == 0) { dst.qword[i] ← combined_lo; } else if (shift_count < 64) { dst.qword[i] ← (combined_lo >> (64 - shift_count)) | (combined_hi << shift_count); } else { dst.qword[i] ← combined_hi << (shift_count - 64); } } }

Example

VPSHLDQ zmm1, zmm2, zmm3/m512, 3

Encoding

Binary Layout

EVEX

Format EVEX

Opcode EVEX.512.66.0F3A.W1 71 /r /ib

Extension AVX-512-VBMI2

Operands

dest
512-bit ZMM AVX-512 register
src1
512-bit ZMM AVX-512 register
src2
512-bit ZMM AVX-512 register or Memory operand
src3
8-bit signed immediate

Reference (Intel® SDM)

Instruction Forms

Opcode	Instruction	Op/En	64/32-bit Mode	CPUID	Description
EVEX.128.66.0F3A.W1 70 /r /ib	VPSHLDW xmm1{k1}{z}, xmm2, xmm3/m128, imm8	A	V/V	(AVX512_VBMI2 AND AVX512VL) OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into xmm1.
EVEX.256.66.0F3A.W1 70 /r /ib	VPSHLDW ymm1{k1}{z}, ymm2, ymm3/m256, imm8	A	V/V	(AVX512_VBMI2 AND AVX512VL) OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into ymm1.
EVEX.512.66.0F3A.W1 70 /r /ib	VPSHLDW zmm1{k1}{z}, zmm2, zmm3/m512, imm8	A	V/V	AVX512_VBMI2 OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into zmm1.
EVEX.128.66.0F3A.W0 71 /r /ib	VPSHLDD xmm1{k1}{z}, xmm2, xmm3/m128/m32bcst, imm8	B	V/V	(AVX512_VBMI2 AND AVX512VL) OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into xmm1.
EVEX.256.66.0F3A.W0 71 /r /ib	VPSHLDD ymm1{k1}{z}, ymm2, ymm3/m256/m32bcst, imm8	B	V/V	(AVX512_VBMI2 AND AVX512VL) OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into ymm1.
EVEX.512.66.0F3A.W0 71 /r /ib	VPSHLDD zmm1{k1}{z}, zmm2, zmm3/m512/m32bcst, imm8	B	V/V	AVX512_VBMI2 OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into zmm1.
EVEX.128.66.0F3A.W1 71 /r /ib	VPSHLDQ xmm1{k1}{z}, xmm2, xmm3/m128/m64bcst, imm8	B	V/V	(AVX512_VBMI2 AND AVX512VL) OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into xmm1.
EVEX.256.66.0F3A.W1 71 /r /ib	VPSHLDQ ymm1{k1}{z}, ymm2, ymm3/m256/m64bcst, imm8	B	V/V	(AVX512_VBMI2 AND AVX512VL) OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into ymm1.
EVEX.512.66.0F3A.W1 71 /r /ib	VPSHLDQ zmm1{k1}{z}, zmm2, zmm3/m512/m64bcst, imm8	B	V/V	AVX512_VBMI2 OR AVX10.1	Concatenate destination and source operands, extract result shifted to the left by constant value in imm8 into zmm1.

Instruction Operand Encoding

Op/En	Tuple Type	Operand 1	Operand 2	Operand 3	Operand 4
A	Full Mem	ModRM:reg (w)	EVEX.vvvv (r)	ModRM:r/m (r)	imm8 (r)
B	Full	ModRM:reg (w)	EVEX.vvvv (r)	ModRM:r/m (r)	imm8 (r)

Description

Concatenate packed data, extract result shifted to the left by constant value. This instruction supports memory fault suppression. VPSHLD—Concatenate and Shift Packed Data Left Logical Vol. 2C 5-622

Operation

VPSHLDW DEST, SRC2, SRC3, imm8
(KL, VL) = (8, 128), (16, 256), (32, 512)
FOR j := 0 TO KL-1:
IF MaskBit(j) OR *no writemask*:
tmp := concat(SRC2.word[j], SRC3.word[j]) << (imm8 & 15)
DEST.word[j] := tmp.word[1]
ELSE IF *zeroing*:
DEST.word[j] := 0
*ELSE DEST.word[j] remains unchanged*
DEST[MAX_VL-1:VL] := 0

VPSHLDD DEST, SRC2, SRC3, imm8
(KL, VL) = (4, 128), (8, 256), (16, 512)
FOR j := 0 TO KL-1:
IF SRC3 is broadcast memop:
tsrc3 := SRC3.dword[0]
ELSE:
tsrc3 := SRC3.dword[j]
IF MaskBit(j) OR *no writemask*:
tmp := concat(SRC2.dword[j], tsrc3) << (imm8 & 31)
DEST.dword[j] := tmp.dword[1]
ELSE IF *zeroing*:
DEST.dword[j] := 0
*ELSE DEST.dword[j] remains unchanged*
DEST[MAX_VL-1:VL] := 0

VPSHLDQ DEST, SRC2, SRC3, imm8
(KL, VL) = (2, 128), (4, 256), (8, 512)
FOR j := 0 TO KL-1:
IF SRC3 is broadcast memop:
tsrc3 := SRC3.qword[0]
ELSE:
tsrc3 := SRC3.qword[j]
IF MaskBit(j) OR *no writemask*:
tmp := concat(SRC2.qword[j], tsrc3) << (imm8 & 63)
DEST.qword[j] := tmp.qword[1]
ELSE IF *zeroing*:
DEST.qword[j] := 0
*ELSE DEST.qword[j] remains unchanged*
DEST[MAX_VL-1:VL] := 0





VPSHLD—Concatenate and Shift Packed Data Left Logical                                                                                      Vol. 2C 5-623

Intel C/C++ Compiler Intrinsic Equivalent

VPSHLDD __m128i _mm_shldi_epi32(__m128i, __m128i, int);
VPSHLDD __m128i _mm_mask_shldi_epi32(__m128i, __mmask8, __m128i, __m128i, int);
VPSHLDD __m128i _mm_maskz_shldi_epi32(__mmask8, __m128i, __m128i, int);
VPSHLDD __m256i _mm256_shldi_epi32(__m256i, __m256i, int);
VPSHLDD __m256i _mm256_mask_shldi_epi32(__m256i, __mmask8, __m256i, __m256i, int);
VPSHLDD __m256i _mm256_maskz_shldi_epi32(__mmask8, __m256i, __m256i, int);
VPSHLDD __m512i _mm512_shldi_epi32(__m512i, __m512i, int);
VPSHLDD __m512i _mm512_mask_shldi_epi32(__m512i, __mmask16, __m512i, __m512i, int);
VPSHLDD __m512i _mm512_maskz_shldi_epi32(__mmask16, __m512i, __m512i, int);
VPSHLDQ __m128i _mm_shldi_epi64(__m128i, __m128i, int);
VPSHLDQ __m128i _mm_mask_shldi_epi64(__m128i, __mmask8, __m128i, __m128i, int);
VPSHLDQ __m128i _mm_maskz_shldi_epi64(__mmask8, __m128i, __m128i, int);
VPSHLDQ __m256i _mm256_shldi_epi64(__m256i, __m256i, int);
VPSHLDQ __m256i _mm256_mask_shldi_epi64(__m256i, __mmask8, __m256i, __m256i, int);
VPSHLDQ __m256i _mm256_maskz_shldi_epi64(__mmask8, __m256i, __m256i, int);
VPSHLDQ __m512i _mm512_shldi_epi64(__m512i, __m512i, int);
VPSHLDQ __m512i _mm512_mask_shldi_epi64(__m512i, __mmask8, __m512i, __m512i, int);
VPSHLDQ __m512i _mm512_maskz_shldi_epi64(__mmask8, __m512i, __m512i, int);
VPSHLDW __m128i _mm_shldi_epi16(__m128i, __m128i, int);
VPSHLDW __m128i _mm_mask_shldi_epi16(__m128i, __mmask8, __m128i, __m128i, int);
VPSHLDW __m128i _mm_maskz_shldi_epi16(__mmask8, __m128i, __m128i, int);
VPSHLDW __m256i _mm256_shldi_epi16(__m256i, __m256i, int);
VPSHLDW __m256i _mm256_mask_shldi_epi16(__m256i, __mmask16, __m256i, __m256i, int);
VPSHLDW __m256i _mm256_maskz_shldi_epi16(__mmask16, __m256i, __m256i, int);
VPSHLDW __m512i _mm512_shldi_epi16(__m512i, __m512i, int);
VPSHLDW __m512i _mm512_mask_shldi_epi16(__m512i, __mmask32, __m512i, __m512i, int);
VPSHLDW __m512i _mm512_maskz_shldi_epi16(__mmask32, __m512i, __m512i, int);

Exceptions

SIMD Floating-Point Exceptions

None.

Other Exceptions

See Table 2-51, “Type E4 Class Exception Conditions.” VPSHLD—Concatenate and Shift Packed Data Left Logical Vol. 2C 5-624