pub fn _mm256_reduce_max_epu16(a: __m256i) -> u16
Reduce the packed unsigned 16-bit integers in a by maximum. Returns the maximum of all elements in a.
Intel’s documentation