Function _mm512_insertf32x4

#[target_feature(enable = "avx512f")]
pub const fn _mm512_insertf32x4(a: __m512, b: __m128, IMM8: i32) -> __m512

Copy a to dst, then insert 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from b into dst at the location specified by imm8.

Intel's documentation