This is supported on x86-64 and target feature
avx512f,avx512vl
only.Expand description
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.