rcpss
Reciprocal Scalar Single-Precision
RCPSS xmm1, xmm2/m32
Computes approximate reciprocal (1/x) of low float.
Details
Computes an approximate reciprocal (1/x) of the low 32-bit single-precision floating-point value in the source operand and stores the result in the low 32 bits of the destination XMM register, while preserving the upper 96 bits of xmm1. No CPU flags are affected. The result has a maximum relative error of approximately 1.5×2^-12.
Pseudocode Operation
xmm1[0:31] ← 1.0 / src[0:31]; xmm1[32:127] ← xmm1[32:127];
Example
RCPSS xmm1, xmm2/m32
Encoding
Binary Layout
F3
+0
0F
+1
53
+2
Operands
-
dest
128-bit XMM SIMD register -
src
128-bit XMM SIMD register or Memory operand
Reference (Intel® SDM)
Instruction Forms
| Opcode | Instruction | Op/En | 64/32-bit Mode | CPUID | Description |
|---|---|---|---|---|---|
| F3 0F 53 /r | RCPSS xmm1, xmm2/m32 | RM | V/V | SSE | Computes the approximate reciprocal of the scalar single precision floating-point value in xmm2/m32 and stores the result in xmm1. |
| VEX.LIG.F3.0F.WIG 53 /r | VRCPSS xmm1, xmm2, xmm3/m32 | AVX | RVM V/V Computes the approximate reciprocal of the scalar single precision floating-point value in xmm3/m32 and stores the result in xmm1. Also, upper single precision floatingpoint values (bits[127:32]) from xmm2 are copied to xmm1[127:32]. |
Description
Computes of an approximate reciprocal of the low single precision floating-point value in the source operand (second operand) and stores the single precision floating-point result in the destination operand. The source operand can be an XMM register or a 32-bit memory location. The destination operand is an XMM register. The three high-order doublewords of the destination operand remain unchanged. See Figure 10-6 in the Intel® 64 and
IA-32 Architectures Software Developer’s Manual, Volume 1, for an illustration of a scalar single precision floatingpoint operation.
The relative error for this approximation is:
|Relative Error| ≤ 1.5 ∗ 2−12
The RCPSS instruction is not affected by the rounding control bits in the MXCSR register. When a source value is a
0.0, an ∞ of the sign of the source value is returned. A denormal source value is treated as a 0.0 (of the same sign).
Tiny results (see Section 4.9.1.5, “Numeric Underflow Exception (#U)” in Intel® 64 and IA-32 Architectures Software Developer’s Manual, Volume 1) are always flushed to 0.0, with the sign of the operand. (Input values greater than or equal to |1.11111111110100000000000B∗2125| are guaranteed to not produce tiny results; input values less than or equal to |1.00000000000110000000001B*2126| are guaranteed to produce tiny results, which are in turn flushed to 0.0; and input values in between this range may or may not produce tiny results, depending on the implementation.) When a source value is an SNaN or QNaN, the SNaN is converted to a QNaN or the source QNaN is returned.
In 64-bit mode, using a REX prefix in the form of REX.R permits this instruction to access additional registers (XMM8-XMM15).
128-bit Legacy SSE version: The first source operand and the destination operand are the same. Bits (MAXVL1:32) of the corresponding YMM destination register remain unchanged.
VEX.128 encoded version: Bits (MAXVL-1:128) of the destination YMM register are zeroed.
Operation
RCPSS (128-bit Legacy SSE Version) DEST[31:0] := APPROXIMATE(1/SRC[31:0]) DEST[MAXVL-1:32] (Unmodified) RCPSS—Compute Reciprocal of Scalar Single Precision Floating-Point Values Vol. 2B 4-540 VRCPSS (VEX.128 Encoded Version) DEST[31:0] := APPROXIMATE(1/SRC2[31:0]) DEST[127:32] := SRC1[127:32] DEST[MAXVL-1:128] := 0
Intel C/C++ Compiler Intrinsic Equivalent
RCPSS __m128 _mm_rcp_ss(__m128 a)
Exceptions
SIMD Floating-Point Exceptions
None.
Other Exceptions
See Table 2-22, “Type 5 Class Exception Conditions.”
RCPSS—Compute Reciprocal of Scalar Single Precision Floating-Point Values Vol. 2B 4-541