Abstract

Modern graphics cards take role of powerful computation hardware. This hardware becomes more popular due to purchasing costs and its availability. The advantages of Graphics Processor Unit (GPU) in parallel computation of Self-Organizing Network are described in this paper including a comparison with multi-threaded CPU. The parallelism on GPU is explained in a separated section. Mentioned section is divided into parts with respect to different forms of parallelism. The results of experiments at the end confirmed, that the utilization of GPU brings significant improvements in time of computation in case of large data sets.

Keywords

Weight Vector Graphic Processing Unit Input Vector Shared Memory Graphic Hardware 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Notes

Acknowledgments

This work was supported by the Ministry of Industry and Trade of the Czech Republic, under the grant no. FR-TI1/420.

References

  1. 1.
    Andrecut, M.: Parallel GPU implementation of iterative PCA algorithms 2008. J. Comput. Biol. 16(11), 1593–1599 (2009)Google Scholar
  2. 2.
    Flexer A.: On the use of self-organizing maps for clustering and visualization. In: Principles of Data Mining and Knowledge Discovery 1999. pp. 80–88Google Scholar
  3. 3.
    Hager, G., Zeiser, T., Wellein, G.: Data access optimizations for highly threaded multi-core CPUs with multiple memory controllers. In: Workshop on Large-Scale Parallel Processing 2008 (IPDPS2008), Miami, 18 April 2008Google Scholar
  4. 4.
    Kirk, D.B. Hwu, W.W.: Programming Massively Parallel Processors A Hands-on Approach 2010. ISBN: 978-0-12-381472-2Google Scholar
  5. 5.
    Kohonen, T.: Self-Organizing Maps. Springer, Berlin (1995)CrossRefGoogle Scholar
  6. 6.
    Lee, S., Min, S-j., Eigenmann, R.: OpenMP to GPGPU: a compiler framework for automatic translation and optimization. In: Proceedings of the 14th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, Raleigh, 2009. pp. 101–110Google Scholar
  7. 7.
    Mann, R., Haykin, S.: A parallel implementation of Kohonen’s feature maps on the warp systolic computer. In: Proceedings of the International Joint Conference on Neural Networks (IJCNN-90), Washington, DC 1990. Vol. II, pp. 84–87Google Scholar
  8. 8.
    Myklebust, G., Solheim, J.G., Steen, E.: Wavefront implementation of self-organizing maps on RENNS. In: International Conference on Digital Signal Processing 1995. pp. 268–273Google Scholar
  9. 9.
    Nordström, T.: Designing parallel computers for self organizing maps. Forth Swedish Workshop on Computer System Architecture, Linköping (1992)Google Scholar
  10. 10.
  11. 11.
    nVIDIA®: Cuda Zone. http://www.nvidia.com/object/cuda_home_new.html (2010). Jan 2010
  12. 12.
    OpenMP: API specification for parallel programming, http://openmp.org/wp/, (2010). Jan 2010
  13. 13.
    Openshaw, S.: Turton I.:A parallel Kohonen algorithm for the classification of large spatial datasets. Comput. Geosci. 22, 1019–1026 (1996)CrossRefGoogle Scholar
  14. 14.
    Patnaik, D., Ponce, S.P., Cao, Y., Ramakrishnan, N.: Accelerator-oriented algorithm transformation for temporal data mining. CoRR, abs/0905.2203, (2009)Google Scholar
  15. 15.
    Preis, T., Virnau, P., Paul, W., Schneider, J.J.: Accelerated fluctuation analysis by graphic cards and complex pattern formation in financial markets. New J. Phys. 11(9), 093024 (21pp) (2009)Google Scholar
  16. 16.
    Valova, I., MacLean, D., Beaton, D.: Identification of patterns via region-growing parallel SOM neural network. In: Seventh International Conference on Machine Learning and Applications, ICMLA 2008. pp. 853–858Google Scholar
  17. 17.
    Valova, I., Szer, D., Gueorguieva, N., Buer, A.: A parallel growing architecture for self-organizing maps with unsupervised learning. Neurocomputing 68, 177–195 (2005)Google Scholar
  18. 18.
    Weigang, L.: A study of parallel self-organizing map. In: Proceedings of the International Joint Conference on Neural Networks, Washington, DC 1999Google Scholar
  19. 19.
    Wu, C.H., Hodges, R.E., Wang, C.J.: Parallelizing the self-organization feature map on multiprocessor systems. Parallel Comput. 17(6–7), 821–832 (1991)MATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  1. 1.Department of Computer ScienceFEECS VSB-Technical University of OstravaOstrava PorubaCzech Republic

Personalised recommendations