Convert the unsigned integer value x to a single-precision floating point value in round-towards-zero mode.
x