pub unsafe fn _mm256_sub_epi64(a: __m256i, b: __m256i) -> __m256i
avx2
Subtract packed 64-bit integers in b from packed 64-bit integers in a
b
a
Intel’s documentation