Approximating \(x^ n\) efficiently (Q1327316)

The authors propose a method for a fast approximation of \(x^ n\) for \(0.0< x< 1.0\) to an accuracy \(1/N\), very useful in scientific calculations. The method uses three techniques: domain compaction (very useful in hardware implementations); errors in look-up tables and interpolation; \(\sqrt n\) divide and conquer mechanism. Numerical examples are given for evaluating color light intensities computed for graphics applications with the authors' algorithm and compared with pow routine provided by the standard C Math Library. Consideration for ROM (read only memory) chip implementation of this algorithm were made.