Adaptation of the direct simulation Monte Carlo method for computing on GPU

Authors

  • A.V. Kashkovsky S.A. Khristianovich Institute of Theoretical and Applied Mechanics of SB RAS https://orcid.org/0000-0001-7169-5602
  • A.A. Shershnev S.A. Khristianovich Institute of Theoretical and Applied Mechanics of SB RAS
  • M.S. Ivanov S.A. Khristianovich Institute of Theoretical and Applied Mechanics of SB RAS

Keywords:

CUDA, DSMC, numerical algorithms, graphics processor units

Abstract

An adaptation technique and algorithms of software implementation of the direct simulation Monte Carlo method for a rarefied gas flow on parallel graphical processors are discussed. Some specific features of graphic processors are illustrated. Several recommendations on writing GPU codes are given.

Author Biographies

A.V. Kashkovsky

S.A. Khristianovich Institute of Theoretical and Applied Mechanics of SB RAS
• Researcher

A.A. Shershnev

S.A. Khristianovich Institute of Theoretical and Applied Mechanics of SB RAS
• Junior Researcher

M.S. Ivanov

S.A. Khristianovich Institute of Theoretical and Applied Mechanics of SB RAS
• Head of Laboratory

References

  1. MPI: A message passing interface standard. Knoxville: University of Tennessee, 1994.
  2. NVIDIA GPU Computing Documentation (http://developer.nvidia.com/nvidia-GPU-computing-documentation).
  3. Фролов В. Введение в технологию CUDA // Компьютерная графика и мультимедиа. 6, № 1. 2008 // (http://cgm.computergraphics.ru/issues/issue16/cuda).
  4. Боресков А. Основы CUDA (http://www.steps3d.narod.ru/tutorials/cuda-tutorial.html).
  5. Бёрд Г. Молекулярная газовая динамика. М.: Мир, 1981.
  6. Bird G.A. Molecular gas dynamics and the direct simulation of gas flows. Oxford: Clarendon Press, 1994.
  7. Ivanov M.S., Rogazinsky S.V. Analysis of the numerical techniques of the direct simulation Monte-Carlo method in the rarefied gas dynamics // Soviet J. Numer. Anal. Math. Modelling. 1988. 3, N 6. 453-465.
  8. Matsumoto M., Nishimura T. Mersenne twister: a 623-dimensionally equidistributed uniform pseudorandom number generator // ACM Trans. on Modeling and Computer Simulations. 1998. 8, № 1. 3-30.
  9. CUDA C/C++ SDK CODE Samples (http://developer.nvidia.com/cuda-cc-sdk-code-samples#MersenneTwister).
  10. Kashkovsky A.V., Markelov G.N., Ivanov M.S. An object-oriented software design for the direct simulation Monte Carlo method // AIAA Paper N 2001-2895. 2001.
  11. Kashkovsky A.V., Bondar Ye.A., Zhukova G.A., Ivanov M.S., Gimelshein S.F. Object-oriented software design of real gas effects for the DSMC method // Proc. of the 24th Int. Symp. on Rarefied Gas Dynamics.
  12. Kashkovsky A.V., Vashchenkov P.V., Ivanov M.S. Object-oriented software design for the three-dimensional direct simulation Monte Carlo method // Proc. of the 25th Int. Symp. on Rarefied Gas Dynamics. St. Petersburg, 2006. 456-461.
  13. Kashkovsky A.V., Vashchenkov P.V., Ivanov M.S. Object-oriented software design approach for multidimensional application of direct simulation Monte Carlo method // Proc. of the 13th Int. Conf. on Methods of Aerophysical Research. 5-10 February, 2007. Part 4. Novosibirsk, 2007. 49-54.
New computational technologies

Downloads

Published

28-05-2012

How to Cite

Кашковский А., Шершнёв А., Иванов М. Adaptation of the Direct Simulation Monte Carlo Method for Computing on GPU // Numerical Methods and Programming (Vychislitel’nye Metody i Programmirovanie). 2012. 13. 53-64

Issue

Vol. 13 (2012): Issue 2.

Section

Section 2. Programming