This content has not been validated; it was AI-generated following the x86-64 ISA Documentation generation process to maintain high quality. However, as no human has verified this data, errors may exist. If the accuracy of this data is critical, please consult the Intel SDM. The model used to generate this is gemma4:31b

VRCP14SS

Computes the reciprocal approximation of the lower 32 bits of the source operand with a relative error of less than $2^{-14}$. The result is stored in the destination operand.

The table after the description covers what the source and destinations can be.

DO NOT support LOCK

This instruction is available in 64-bit mode and 32-bit mode. It requires the SSE4.1 instruction set extension.

The instruction does not provide a precise reciprocal; it provides an approximation. To obtain a more accurate result, the programmer SHALL use the result as an initial guess for a Newton-Raphson iteration.

The instruction does not trigger #D, #O, #U, or #P exceptions. If the source operand is $\pm0$, the result is $\pm\infty$. If the source operand is $\pm\infty$ or NaN, the result is $0$ or NaN respectively.