Vectorization of the beam propagation method using cuda technology
V.A. Alekseeev, D.L. Golovashkin

Image Processing Systems Institute of the RAS,
Samara State Aerospace University

Full text of article: Russian language.

Abstract:
We propose a vector algorithm for solving the helmholtz equation through BPM (beam propagation method) by representing the computation in the SIMD (single instruction, multiple data) model. The implementation of the given algorithm on graphic processor NVIDIA GeForce GTS 250 on technology CUDA has shown acceleration of calculations in 22.5 times in comparison with the computation on the central processor Intel Core Duo E7500.

Key words:
helmholtz equation, beam propagation method, vector algorithm, graphics processor, componentwise product, cuda.

References:

  1. Feit, M.D. Light Propagation in Graded-Index Optical Fibers / M.D. Feit, J.A. Fleck // Applied Optics. – 1978. – Vol. 17, N. 24. – P. 3990-3998.
  2. Gerdes, J. Beam-propagation algorithm based on the method of lines / J. Gerdes, R. Pregla // Journal of the Optical Society of America B. – 1991. – Vol. 8, N. 2. – P. 389-394.
  3. Huang, W. The Finite-Difference Vector Beam Propa­ga­tion Method: Analysis and Assessment / W. Huang, C. Xu, S.-T. Chu, S.K. Chaudhuri // Journal of Lightwave Technology. – 1992. – Vol. 10, N. 3. – P. 295-305.
  4. Lu, Y.Y. Some Techniques for Computing Wave Propagation in Optical Waveguides / Y.Y. Lu // Communications in Computational Physics. – 2006. – Vol. 1, N. 6. – P. 1056-1075.
  5. Gavrilov, A.V. The modified method of beam propagation and its application to distribution calculation in wave guides with a changing profile of an indicator of refraction / A.V. Gavrilov // Computer Optics. – 2008. – V. 32, N 1. – P. 15-22.
  6. OlympIOs: design, simulation and mask layout platform //URL: http://www.c2v.nl/products/software/olympios-software.shtml.
  7. RSoft: Photonic Component Design Suite // URL: http://www.rsoftdesign.com/uicontrols/products/brochures/RSoftProductCatalog.pdf.
  8. Golub, G. Matrix Computations / G. Golub, C. Van Loan – Мoscow: “Mir” publisher, 1999. – 548 p. – (in Russian).
  9. Ortega J. Introduction to parallel and vector solution of linear systems / J. Ortega – Moscow: “Mir” publisher, 1991. – 364 p. – (in Russian).
  10. CUDA, Published by NVIDIA Corporation – 2701 San Tomas Expressway Santa Clara, CA 95050,
    http://www.nvidia.com/object/cuda_home_new.html
  11. Evstigneev, N.M. Integration of the equation of Puassona with use of the graphic processor of technology // Computing methods and programming. – 2009. – Т. 10, N 2. – P. 82-87. – (in Russian).
  12. Price, D.K. GPU-based accelerated 2D and 3D FDTD solvers / D.K. Price, J.R. Humphrey, and E.J. Kelmelis //– Physics and Simulation of Optoelectronic Devices XV, vol. 6468 of Proceedings of SPIE, San Jose, Calif, USA, January 2007.
  13. Adinetz, A. Implementing Classical Ray Tracing on GPU – a Case Study of GPU Programming / A. Adinetz, S. Be­re­zin – Proceedings of Graphicon, 2006.
  14. Samarskii, A.A. Introduction in numerical methods /А.А. Sa­marskii – Мoscow: “Nayka” publisher, 1987. – 286 p. – (in Russian).
  15. Kosarev, V.I. 12 lectures on calculus mathematics. An introduction course / V.I. Kosarev – Moscow: “Fizmatlit” publisher, 2000. – 224 p. – (in Russian).

© 2009, ИСОИ РАН
Россия, 443001, Самара, ул. Молодогвардейская, 151; электронная почта: ko@smr.ru ; тел: +7 (846 2) 332-56-22, факс: +7 (846 2) 332-56-20