Function _mm256_zextph128_ph256

#[target_feature(enable = "avx512fp16")]
pub const fn _mm256_zextph128_ph256(a: __m128h) -> __m256h

Cast vector of type __m256h to type __m128h. The upper 8 elements of the result are zeroed. This intrinsic can generate the vzeroupper instruction, but most of the time it does not generate any instructions.

Intel's documentation