Fast floating-point to integer conversion

Input signals in floating-point format are converted into output signals in integer format while handling saturation cases appropriately. Saturation exists when the value of an input signal is outside the output range. Saturation cases are detected by invoking an integer comparison instruction using signals in floating-point format as inputs to tile integer comparison instruction. When implemented in software on processors such as the Intel TM Pentium TM processor, using an integer comparison instruction to compare signals in floating-point format provides more efficient signal conversion than conventional schemes that use a floating-point comparison instruction to compare floating-point signals.