# [−][src]Function core::arch::x86_64::_mm_sad_pu8

pub unsafe fn _mm_sad_pu8(a: __m64, b: __m64) -> __m64

This is supported on

**x86-64 and target feature**only.`sse,mmx`

Subtracts the corresponding 8-bit unsigned integer values of the two
64-bit vector operands and computes the absolute value for each of the
difference. Then sum of the 8 absolute differences is written to the
bits `[15:0]`

of the destination; the remaining bits `[63:16]`

are cleared.