| Opcode/Instruction | Op/En | 64/32 bit Mode Support | CPUID Feature Flag | Description |
|---|---|---|---|---|
| F3 0F 51 /r SQRTSS xmm1, xmm2/m32 | A | V/V | SSE | Computes square root of the low single-precision floating-point value in xmm2/m32 and stores the results in xmm1. |
| VEX.LIG.F3.0F.WIG 51 /r VSQRTSS xmm1, xmm2, xmm3/m32 | B | V/V | AVX | Computes square root of the low single-precision floating-point value in xmm3/m32 and stores the results in xmm1. Also, upper single-precision floating-point values (bits[127:32]) from xmm2 are copied to xmm1[127:32]. |
| EVEX.LIG.F3.0F.W0 51 /r VSQRTSS xmm1 {k1}{z}, xmm2, xmm3/m32{er} | C | V/V | AVX512F | Computes square root of the low single-precision floating-point value in xmm3/m32 and stores the results in xmm1 under writemask k1. Also, upper single-precision floating-point values (bits[127:32]) from xmm2 are copied to xmm1[127:32]. |
| Op/En | Tuple Type | Operand 1 | Operand 2 | Operand 3 | Operand 4 |
| A | NA | ModRM:reg (w) | ModRM:r/m (r) | NA | NA |
| B | NA | ModRM:reg (w) | VEX.vvvv (r) | ModRM:r/m (r) | NA |
| C | Tuple1 Scalar | ModRM:reg (w) | EVEX.vvvv (r) | ModRM:r/m (r) | NA |
Computes the square root of the low single-precision floating-point value in the second source operand and stores the single-precision floating-point result in the destination operand. The second source operand can be an XMM register or a 32-bit memory location. The first source and destination operands is an XMM register.
128-bit Legacy SSE version: The first source operand and the destination operand are the same. Bits (MAXVL-1:32) of the corresponding YMM destination register remain unchanged.
VEX.128 and EVEX encoded versions: Bits 127:32 of the destination operand are copied from the corresponding bits of the first source operand. Bits (MAXVL-1:128) of the destination ZMM register are zeroed.
EVEX encoded version: The low doubleword element of the destination operand is updated according to the writemask.
Software should ensure VSQRTSS is encoded with VEX.L=0. Encoding VSQRTSS with VEX.L=1 may encounter unpredictable behavior across different processor generations.
IF (EVEX.b = 1) AND (SRC2 *is register*)
THEN
SET_RM(EVEX.RC);
ELSE
SET_RM(MXCSR.RM);
FI;
IF k1[0] or *no writemask*
THEN DEST[31:0] ← SQRT(SRC2[31:0])
ELSE
IF *merging-masking* ; merging-masking
THEN *DEST[31:0] remains unchanged*
ELSE ; zeroing-masking
DEST[31:0] ← 0
FI;
FI;
DEST[127:31] ← SRC1[127:31]
DEST[MAXVL-1:128] ← 0
DEST[31:0] ←SQRT(SRC2[31:0]) DEST[127:32] ←SRC1[127:32] DEST[MAXVL-1:128] ←0
DEST[31:0] ←SQRT(SRC2[31:0]) DEST[MAXVL-1:32] (Unmodified)
VSQRTSS __m128 _mm_sqrt_round_ss(__m128 a, __m128 b, int r);
VSQRTSS __m128 _mm_mask_sqrt_round_ss(__m128 s, __mmask8 k, __m128 a, __m128 b, int r);
VSQRTSS __m128 _mm_maskz_sqrt_round_ss( __mmask8 k, __m128 a, __m128 b, int r);
SQRTSS __m128 _mm_sqrt_ss(__m128 a)
Invalid, Precision, Denormal
Non-EVEX-encoded instruction, see Exceptions Type 3.
EVEX-encoded instruction, see Exceptions Type E3.