An integer based square root algorithm
- 66 Downloads
We propose a fast integer based method for computing square roots of floating point numbers. This implies high accuracy and robustness, since no precision will be lost during the computation. Only integer addition and shifts are necessary to obtain the square root. Comparisons made with the modified Newton method indicate that the suggested method is twice as fast for computing floating point square roots.
AMS categories65D15 68C05
Unable to display preview. Download preview PDF.
- 1.T. C. Chen:Automatic computation of exponentials, logarithms, ratios and square roots. IBM J. Res. Dev., 16 (4): 380–388, July 1972.Google Scholar
- 2.W. J. Cody, Jr. and W. Waite:Software Manual for the Elementary Functions. Series in Computational Mathematics. Prentice-Hall, Englewood Cliffs, NJ, 1980.Google Scholar
- 3.V. G. Oklobdzija and M. D. Ercegovac:An on-line square root algorithm. IEEE Trans. Comput., C-31 (1): 70–75, Jan. 1982.Google Scholar
- 4.Stevenson et al.:IEEE Standard for Binary Floating-Point Arithmetic. IEEE Computer Society, 1985. ANSI/IEEE Std. 754–1985.Google Scholar
- 5.J. S. Walther:A unified algorithm for elementary functions. In Proc. AFIPS 1971 Spring Joint Computer Conference, pages 379–385, May 1971.Google Scholar