pub unsafe fn _mm256_hadd_epi32(a: __m256i, b: __m256i) -> __m256i
avx2
Horizontally adds adjacent pairs of 32-bit integers in a and b.
a
b
Intel’s documentation