CHOOSING BETWEEN FIXED- AND FLOATING-POINT IMPLEMENTATIONS CAN BE DIFFICULT. A COMPARISON OF ONE OF EACH TYPE OF PROCESSOR AND SOME SAMPLE APPLICATIONS HIGHLIGHT SOME OF THE PROS AND CONS OF EACH APPROACH. Without a doubt, many algorithms can benefit from employing a floating-point implementation. The code can be simpler and take fewer cycles to execute than fixed-point implementations. However, these benefits do not really matter that much to end designers. Fewer designers are coding in assembly language, because compilers are getting better; coding fixed- or floating-point implementations in C takes similar-complexity. Designers can write floating-point code for a fixed-point processor: in C, but doing so causes a significant performance hit. What matters to end designers is the final system's performance, as well as the Cost and time to market.
展开▼
