Compare the lower single-precision (32-bit) floating-point element in a and b for less-than, and return the boolean result (0 or 1).