pub fn _kadd_mask16(a: __mmask16, b: __mmask16) -> __mmask16
avx512dq
Add 16-bit masks a and b, and store the result in dst.
Intel’s Documentation