Numerical integration of Poisson's equation using a graphics processing unit with CUDA-technology

Authors

  • N.M. Evstigneev

Keywords:

Poisson's equation
cyclic multigrid method
parallel computing
graphics processing unit
CUDA technology

Abstract

A parallel implementation of the cyclic multigrid method for solving a boundary value problem for Poisson's equation in R3 is discussed for a graphics processor unit using NVIDIA CUDA technology. The results obtained by the graphics processing unit is compared with the analytical solution for the Dirichlet problem and with the numerical CPU-solution for the Dirichlet-Neumann problem. The acceleration of the parallel NVIDIA GeForce 8800 GTX code compared to the AMD Athlon 64X2 4800+ serial code is found to be about 200 times for 1 000 000 discrete elements. Moreover, a 8-core workstation based on two Intel(R) Xeon(R) 2.33HHz CPUs is found to be slower by 40 times than that of the GPU code.

New computational technologies

Downloads

Published

2020-11-10

Issue

Vol. 10 (2009): Issue 2.

Section

Section 1. Numerical methods and applications

Author Biography

N.M. Evstigneev

Institute for Systems Analysis of RAS (ISA RAS)
• Senior Researcher

References

  1. Deuflhard P., Leinen P., Yserentant H. Concepts of an adaptive hierarchical finite element code // Impact Comput. Sci. Eng. 1989. 1. 3-35.
  2. Shaidurov V.V. Some estimates of the rate of convergence for the cascadic conjugate gradient method // Comput. Math. Appl. 1996. 31. 161-171.
  3. Shi Z., Xu X. Cascadic multigrid method for elliptic problems // East-West J. Numer. Math. 1999. 3. 199-209.
  4. Braess D., Dahmen W. A cascadic multigrid algorithm for the Stokes equation // Numer. Math. 1999. 82. 179-191.
  5. Shi Z., Xu X. Cascadic multigrid method for parabolic problems. Preprint, 1999.
  6. Bornemann F.A., Krause R. Classical and cascadic multigrid - a methodical comparison // Proc. of the Ninth International Conference on Domain Decomposition. Bergen: Domain Decomposition Press, 1998. 64-71.
  7. Боярченков А.С., Поташников С.И. Использование графических процессоров и технологии CUDA для задач молекулярной динамики // Вычислительные методы и программирование. 2009. 10, № 1. 13-27.
  8. http://developer.download.nvidia.com/compute/cuda/2_0/docs/NVIDIA_CUDA_Programming_Guide_2.0.pdf
  9. Hyman J.M. Mesh refinement and local inversion of elliptic differential equations // J. Comput. Phys. 1977. 23. 124-134.
  10. Формалев В.Ф., Ревизников Д.Л. Численные методы. М.: Физматлит, 2006.
  11. Degenhardt R., Berz M. High accuracy description of the fringe field in particle spectrographs // Nuclear Instruments and Methods. 1999. A427, 151-156.
  12. http://www.roylongbottom.org.uk/index.htm#anchorNew