Compare packed single-precision (32-bit) floating-point elements in a and b, and return packed maximum values.
See Implementation
Compare packed single-precision (32-bit) floating-point elements in a and b, and return packed maximum values.