pub fn _mm256_mask_reduce_or_epi16(k: u16, a: __m256i) -> i16
Reduce the packed 16-bit integers in a by bitwise OR using mask k. Returns the bitwise OR of all active elements in a.
Intel’s documentation