pub fn _mm512_set1_ph(a: f16) -> __m512h
stdarch_x86_avx512_f16
Broadcast the half-precision (16-bit) floating-point value a to all elements of dst.
Intel’s documentation