Table 1: Performance gains possible through a combination of advanced architectures and optimizing compilers. Improving performance and efficiency, based on 500 input samples, 32 lags, total multiplies of 500×32=16,000, and total loops=total mul/(mul/loop).