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

MPSADBW

Computes the sum of absolute differences between 8-bit signed integers. It treats the source and destination operands as vectors of 8-bit signed integers. For each 32-bit block, it calculates the absolute difference between corresponding bytes of the two operands for four different offsets, sums these differences, and stores the resulting four unsigned 32-bit integers in the destination.

The table below covers the supported source and destination operands.

DO NOT support LOCK

This instruction is only available in 64-bit mode or compatibility mode. It requires the SSE4.1 instruction set extension.

The destination register is overwritten by the result; therefore, if the source xmm register is also used as the destination, the original data is lost. The operation ignores the upper 64 bits of the xmm registers if operating on 128-bit vectors. Alignment of the m16 operand must follow the memory alignment rules for the specific processor architecture to avoid performance penalties or alignment faults.