• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Article

Employing AVX Vectorization to Improve the Performance of Random Number Generators

Programming and Computer Software. 2017. Vol. 43. No. 3. P. 145-160.
Barash L., Guskova M., Shchur L.
Translator: Y. Kornienko.

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 10^19 independent ran-dom number streams. The AVX implementations yield exactly the same pseudorandom sequences as the orig-inal algorithms do, while being up to 40 times faster than the ANSI C implementations.