pub unsafe fn _mm256_adds_epi16(a: __m256i, b: __m256i) -> __m256i
This is supported on x86-64 and target feature avx2 only.
Expand description

Adds packed 16-bit integers in a and b using saturation.

Intel’s documentation