Function core::arch::x86::_mm512_maddubs_epi16

pub unsafe fn _mm512_maddubs_epi16(a: __m512i, b: __m512i) -> __m512i

🔬 This is a nightly-only experimental API. (stdsimd #48556)

Available on x86 and target feature avx512bw only.

Vertically multiply each unsigned 8-bit integer from a with the corresponding signed 8-bit integer from b, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers, and pack the saturated results in dst.

Intel’s documentation