Email this article Employing AVX vectorization to improve the performance of random number generators
- Authors: Barash L.Y.1,2, Guskova M.S.2,3, Shchur L.N.1,2,3
-
Affiliations:
- Landau Institute for Theoretical Physics
- Science Center in Chernogolovka
- National Research University Higher School of Economics
- Issue: Vol 43, No 3 (2017)
- Pages: 145-160
- Section: Article
- URL: https://journals.rcsi.science/0361-7688/article/view/176503
- DOI: https://doi.org/10.1134/S0361768817030033
- ID: 176503
Cite item
Open Access
Access granted
Subscription Access
Abstract
By the example of the RNGAVXLIB random number generator library, this paper considers some approaches to employing AVX vectorization for calculation speedup. The RNGAVXLIB library contains AVX implementations of modern generators and the routines allowing one to initialize up to 1019 independent random number streams. The AVX implementations yield exactly the same pseudorandom sequences as the original algorithms do, while being up to 40 times faster than the ANSI C implementations.
About the authors
L. Yu. Barash
Landau Institute for Theoretical Physics; Science Center in Chernogolovka
Author for correspondence.
Email: barash@itp.ac.ru
Russian Federation, Chernogolovka, 142432; Chernogolovka, 142432
M. S. Guskova
Science Center in Chernogolovka; National Research University Higher School of Economics
Email: barash@itp.ac.ru
Russian Federation, Chernogolovka, 142432; Moscow, 101000
L. N. Shchur
Landau Institute for Theoretical Physics; Science Center in Chernogolovka; National Research University Higher School of Economics
Email: barash@itp.ac.ru
Russian Federation, Chernogolovka, 142432; Chernogolovka, 142432; Moscow, 101000
Supplementary files
