Intel® C++ Compiler 16.0 User and Reference Guide
Loads and broadcasts packed single-precision floating point values. The corresponding Intel® AVX instruction is VBROADCASTF128.
extern __m256 _mm256_broadcast_ps(__m128 const *a);  | 
*a  | 
      pointer to a memory location that can hold constant 128-bit float32 values | 
Loads 128-bit float32 values from the specified address pointed to by a, and broadcasts it to all elements in the destination 256-bit vector.
Result of the load and broadcast operation.