|
Program Systems: Theory and Applications, 2016, Volume 7, Issue 1, Pages 61–97
(Mi ps204)
|
|
|
|
This article is cited in 1 scientific paper (total in 1 paper)
Hardware, software and distributed supercomputer systems
Parallel multiple-precision arithmetic based on residue number system
K. S. Isupov, V. S. Knyazkov Vyatka State University
Abstract:
This paper deals with algorithms of multiple-precision arithmetic, based on the use of multi module residue number systems for representing of arbitrary length significands of floating-point numbers; the exponent is represented in the binary number system. Such number representation provides a large dynamic range and allows for effective parallelization of arithmetic operations on the digits of multiple-precision significands across RNS modules. This agrees well with the architectural features of modern parallel computing systems. Additionally, the attributive information which provides a fast estimation for the relative value of significand and allows you to increase the speed of executing complex non-modular operations in RNS, such as comparison, overflow control, rounding, etc., is included into the number format. Results of an experimental study on precision, performance and SIMD efficiency of multiple-precision algorithms are presented. (In Russian).
Key words and phrases:
computer arithmetic, high-precision computations, parallel algorithms, residue number system, SIMD.
Received: 27.01.2016 Accepted: 24.02.2016
Citation:
K. S. Isupov, V. S. Knyazkov, “Parallel multiple-precision arithmetic based on residue number system”, Program Systems: Theory and Applications, 7:1 (2016), 61–97
Linking options:
https://www.mathnet.ru/eng/ps204 https://www.mathnet.ru/eng/ps/v7/i1/p61
|
Statistics & downloads: |
Abstract page: | 344 | Full-text PDF : | 138 | References: | 39 |
|