x86 Instruction Set Reference
SQRTPD
Compute Square Roots of Packed Double-Precision Floating-Point Values
| Opcode | Mnemonic | Description |
|---|---|---|
66 0F 51 /r |
SQRTPD xmm1, xmm2/m128 |
Compute the square root of the packed doubleprecision floating-point values in xmm2/m128 and store the results in xmm1. |
| Description |
|---|
|
Performs an SIMD computation of the square roots of the two packed double-precision floatingpoint values in the source operand (second operand) stores the packed double-precision floatingpoint results in the destination operand. The source operand can be an XMM register or a 128- bit memory location. The destination operand is an XMM register. See Figure 11-3 in the IA-32 Intel Architecture Software Developer's Manual, Volume 1 for an illustration of an SIMD double-precision floating-point operation. |
| Operation |
|---|
Destination[0..63] = SquareRoot(Source[0..63]); Destination[64..127] = SquareRoot(Source[64..127]); |
| Instruction | Latency | Throughput | Execution Unit |
|---|---|---|---|
CPUID | 0F3n/0F2n/069n | 0F3n/0F2n/069n | 0F2n |
SQRTPD xmm, xmm | 70/69/58+57 | 70/69/114 | FP_DIV |