Unpack and interleave single-precision (32-bit) floating-point elements from the high half a and b.
See Implementation
Unpack and interleave single-precision (32-bit) floating-point elements from the high half a and b.