pub fn _mm256_mask_reduce_add_epi16(k: u16, a: __m256i) -> i16
Reduce the packed 16-bit integers in a by addition using mask k. Returns the sum of all active elements in a.
Intel’s documentation