CVTDQ2PD — Convert Packed Doubleword Integers to Packed Double-Precision Floating-Point Values

Opcode/Instruction Op/En 64/32 bit Mode Support CPUID Feature Flag Description
F3 0F E6 /r CVTDQ2PD xmm1, xmm2/m64 A V/V SSE2 Convert two packed signed doubleword integers from xmm2/mem to two packed double-precision floating-point values in xmm1.
VEX.128.F3.0F.WIG E6 /r VCVTDQ2PD xmm1, xmm2/m64 A V/V AVX Convert two packed signed doubleword integers from xmm2/mem to two packed double-precision floating-point values in xmm1.
VEX.256.F3.0F.WIG E6 /r VCVTDQ2PD ymm1, xmm2/m128 A V/V AVX Convert four packed signed doubleword integers from xmm2/mem to four packed double-precision floating-point values in ymm1.
EVEX.128.F3.0F.W0 E6 /r VCVTDQ2PD xmm1 {k1}{z}, xmm2/m128/m32bcst B V/V AVX512VL AVX512F Convert 2 packed signed doubleword integers from xmm2/m128/m32bcst to eight packed double-precision floating-point values in xmm1 with writemask k1.
EVEX.256.F3.0F.W0 E6 /r VCVTDQ2PD ymm1 {k1}{z}, xmm2/m128/m32bcst B V/V AVX512VL AVX512F Convert 4 packed signed doubleword integers from xmm2/m128/m32bcst to 4 packed double-precision floating-point values in ymm1 with writemask k1.
EVEX.512.F3.0F.W0 E6 /r VCVTDQ2PD zmm1 {k1}{z}, ymm2/m256/m32bcst B V/V AVX512F Convert eight packed signed doubleword integers from ymm2/m256/m32bcst to eight packed double-precision floating-point values in zmm1 with writemask k1.

Instruction Operand Encoding

Op/En Tuple Type Operand 1 Operand 2 Operand 3 Operand 4
A NA ModRM:reg (w) ModRM:r/m (r) NA NA
B Half ModRM:reg (w) ModRM:r/m (r) NA NA

Description

Converts two, four or eight packed signed doubleword integers in the source operand (the second operand) to two, four or eight packed double-precision floating-point values in the destination operand (the first operand).

EVEX encoded versions: The source operand can be a YMM/XMM/XMM (low 64 bits) register, a 256/128/64-bit memory location or a 256/128/64-bit vector broadcasted from a 32-bit memory location. The destination operand is a ZMM/YMM/XMM register conditionally updated with writemask k1. Attempt to encode this instruction with EVEX embedded rounding is ignored.

VEX.256 encoded version: The source operand is an XMM register or 128- bit memory location. The destination operand is a YMM register.

VEX.128 encoded version: The source operand is an XMM register or 64- bit memory location. The destination operand is a XMM register. The upper Bits (MAXVL-1:128) of the corresponding ZMM register destination are zeroed.

128-bit Legacy SSE version: The source operand is an XMM register or 64- bit memory location. The destination operand is an XMM register. The upper Bits (MAXVL-1:128) of the corresponding ZMM register destination are unmodified.

VEX.vvvv and EVEX.vvvv are reserved and must be 1111b, otherwise instructions will #UD.

X3 X2 X1 X0 SRC X1 X0 X3 X2 DEST
Figure 3-11. CVTDQ2PD (VEX.256 encoded version)

Operation

VCVTDQ2PD (EVEX encoded versions) when src operand is a register

(KL, VL) = (2, 128), (4, 256), (8, 512)
FOR j←0 TO KL-1
    i←j * 64
    k←j * 32
    IF k1[j] OR *no writemask*
        THEN DEST[i+63:i]←
            Convert_Integer_To_Double_Precision_Floating_Point(SRC[k+31:k])
        ELSE
            IF *merging-masking* ; merging-masking
                THEN *DEST[i+63:i] remains unchanged*
                ELSE
                        ; zeroing-masking
                    DEST[i+63:i] ← 0
            FI
    FI;
ENDFOR
DEST[MAXVL-1:VL] ← 0

VCVTDQ2PD (EVEX encoded versions) when src operand is a memory source

(KL, VL) = (2, 128), (4, 256), (8, 512)
FOR j←0 TO KL-1
    i←j * 64
    k←j * 32
    IF k1[j] OR *no writemask*
        THEN
            IF (EVEX.b = 1)
                THEN
                    DEST[i+63:i] ←
            Convert_Integer_To_Double_Precision_Floating_Point(SRC[31:0])
                ELSE
                    DEST[i+63:i] ←
            Convert_Integer_To_Double_Precision_Floating_Point(SRC[k+31:k])
            FI;
        ELSE
            IF *merging-masking* ; merging-masking
                THEN *DEST[i+63:i] remains unchanged*
                ELSE ; zeroing-masking
                    DEST[i+63:i] ← 0
            FI
    FI;
ENDFOR
DEST[MAXVL-1:VL] ← 0

VCVTDQ2PD (VEX.256 encoded version)

DEST[63:0] ← Convert_Integer_To_Double_Precision_Floating_Point(SRC[31:0])
DEST[127:64] ← Convert_Integer_To_Double_Precision_Floating_Point(SRC[63:32])
DEST[191:128] ← Convert_Integer_To_Double_Precision_Floating_Point(SRC[95:64])
DEST[255:192] ← Convert_Integer_To_Double_Precision_Floating_Point(SRC[127:96)
DEST[MAXVL-1:256] ← 0

VCVTDQ2PD (VEX.128 encoded version)

DEST[63:0] ← Convert_Integer_To_Double_Precision_Floating_Point(SRC[31:0])
DEST[127:64] ← Convert_Integer_To_Double_Precision_Floating_Point(SRC[63:32])
DEST[MAXVL-1:128] ← 0

CVTDQ2PD (128-bit Legacy SSE version)

DEST[63:0] ← Convert_Integer_To_Double_Precision_Floating_Point(SRC[31:0])
DEST[127:64] ← Convert_Integer_To_Double_Precision_Floating_Point(SRC[63:32])
DEST[MAXVL-1:128] (unmodified)

Intel C/C++ Compiler Intrinsic Equivalent

VCVTDQ2PD __m512d _mm512_cvtepi32_pd( __m256i a);
VCVTDQ2PD __m512d _mm512_mask_cvtepi32_pd( __m512d s, __mmask8 k, __m256i a);
VCVTDQ2PD __m512d _mm512_maskz_cvtepi32_pd( __mmask8 k, __m256i a);
VCVTDQ2PD __m256d _mm256_cvtepi32_pd (__m128i src);
VCVTDQ2PD __m256d _mm256_mask_cvtepi32_pd( __m256d s, __mmask8 k, __m256i a);
VCVTDQ2PD __m256d _mm256_maskz_cvtepi32_pd( __mmask8 k, __m256i a);
VCVTDQ2PD __m128d _mm_mask_cvtepi32_pd( __m128d s, __mmask8 k, __m128i a);
VCVTDQ2PD __m128d _mm_maskz_cvtepi32_pd( __mmask8 k, __m128i a);
CVTDQ2PD __m128d _mm_cvtepi32_pd (__m128i src)

Other Exceptions

VEX-encoded instructions, see Exceptions Type 5;

EVEX-encoded instructions, see Exceptions Type E5.

#UD If VEX.vvvv != 1111B or EVEX.vvvv != 1111B.