Function _mm_set1_ph

#[target_feature(enable = "avx512fp16")]
pub const fn _mm_set1_ph(a: f16) -> __m128h

Broadcast the half-precision (16-bit) floating-point value a to all elements of dst.

Intel's documentation