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

VFMSUBADD132PH

Performs a fused multiply-subtract and add operation on half-precision floating-point values. It computes the result of (a * b) - c + d for each element in the source operands.

The following table describes the supported source and destination types.

DO NOT support LOCK

This instruction requires the AVX-512 FP16 extension. It is NOT available in compatibility mode; it MUST be executed in 64-bit mode.

To avoid precision loss or unexpected behavior, ensure that the destination register is not used as a source for a different operand unless the intended behavior is accumulation. Be aware that this instruction may trigger floating-point exceptions: #O, #U, and #P depending on the rounding mode and the magnitude of the intermediate product. Underflow (#U) and Inexact (#P) flags are set if the rounded result differs from the mathematical result.