Compute the square root of packed single-precision (32-bit) floating-point elements in a.
See Implementation
Compute the square root of packed single-precision (32-bit) floating-point elements in a.