pub fn _mm_mask_reduce_add_epi16(k: u8, a: __m128i) -> 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