Convert the lower packed unsigned 8-bit integers in a to packed single-precision (32-bit) floating-point element.
See Implementation
Convert the lower packed unsigned 8-bit integers in a to packed single-precision (32-bit) floating-point element.