Abstract
We develop a bit manipulation technique for single precision floating point numbers which leads to new algorithms for fast computation of the cube root and inverse cube root. It uses the modified iterative Newton–Raphson method (the first order of convergence) and Householder method (the second order of convergence) to increase the accuracy of the results. The proposed algorithms demonstrate high efficiency and reduce error several times in the first iteration in comparison with known algorithms. After two iterations 22.84 correct bits were obtained for single precision. Experimental tests showed that our novel algorithm is faster and more accurate than library functions for microcontrollers.
Subject
Energy (miscellaneous),Energy Engineering and Power Technology,Renewable Energy, Sustainability and the Environment,Electrical and Electronic Engineering,Control and Optimization,Engineering (miscellaneous)
Reference25 articles.
1. The Mathematical-Function Computation Handbook: Programming Using the Math. CWPortable Software Library;Beebe,2017
2. Efficient Floating-Point Division for Digital Signal Processing Application [Tips & Tricks]
3. https://web.archive.org/web/20150511044204/http://www.daxia.com/bibis/upload/406Fast_Inverse_Square_Root.pdf
4. Fast calculation of inverse square root with the use of magic constant – analytical approach
5. Computing a Real Cube Root. November 18, 1991https://csclub.uwaterloo.ca/~pbarfuss/qbrt.pdf