3-596
INSTRUCTION SET REFERENCE
RCPPS—Packed Single-FP Reciprocal
Description
RCPPS returns an approximation of the reciprocal of the SP FP numbers from xmm2/m128. The
maximum error for this approximation is:
Error <=1.5x2-12
Figure 3-86. Operation of the RCPPS Instruction
Operation
DEST[31-0] = APPROX (1.0/(SRC/m128[31-0]));
DEST[63-32] = APPROX (1.0/(SRC/m128[63-32]));
DEST[95-64] = APPROX (1.0/(SRC/m128[95-64]));
DEST[127-96] = APPROX (1.0/(SRC/m128[127-96]));
Intel C/C++ Compiler Intrinsic Equivalent
__m128 _mm_rcp_ps(__m128 a)
Computes the approximations of the reciprocals of the four SP FP values of a.
OpcodeInstruction
Description
0F,53,/rRCPPS xmm1, xmm2/m128Return a packed approximation of the reciprocal of
XMM2/Mem.
Xmm1
Xmm1
Xmm2/
m128
5.0
125.0
0.2
=
=
=
=
