Advertisement

Towards a Complexity Model for Design and Analysis of PGAS-Based Algorithms

  • Mohamed Bakhouya
  • Jaafar Gaber
  • Tarek El-Ghazawi
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4782)

Abstract

Many new Partitioned Global Address Space (PGAS) programming languages have recently emerged and are becoming ubiquitously available on nearly all modern parallel architectures. PGAS programming languages provide ease-of-use through a global shared address space while emphasizing performance by providing locality awareness and a partition of the address space. Examples of PGAS languages include the Unified Parallel C (UPC), Co-array Fortran, and Titanium languages. Therefore, the interest in complexity design and analysis of PGAS algorithms is growing and a complexity model to capture implicit communication and fine-grain programming style is required. In this paper, a complexity model is developed to characterize the performance of algorithms based on the PGAS programming model. The experimental results shed further light on the impact of data distributions on locality and performance and confirm the accuracy of the complexity model as a useful tool for the design and analysis of PGAS-based algorithms.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    El-Ghazawi, T., Carlson, W., Sterling, T., Yelick, K.: UPC: Distributed Shared Memory Programming. Book. John Wiley and Sons Inc., New York (2005)Google Scholar
  2. 2.
    Gaber, J.: Complexity Measure Approach for Partitioned Shared Memory Model, Application to UPC. Research report RR-10-04. Universite de Technologie de Belfort-Montbeliard (2004)Google Scholar
  3. 3.
    Cantonnet, F., El Ghazawi, T., Lorenz, P., Gaber, J.: Fast Address Translation Techniques for Distributed Shared Memory Compilers. In: International Parallel and Distributed Processing Symposium IPDPS 2006 (2006)Google Scholar
  4. 4.
    Cameron, K.W., Sun, X.-H.: Quantifying Locality Effect in Data Access Delay: Memory logP. In: IPDPS 2003. Proc. of the 17th International Symposium on Parallel and Distributed Processing, p. 48.2 (2003)Google Scholar
  5. 5.
    Zhang, S., Seidel, R.Z.: A performance model for fine-grain accesses in UPC. In: 20th International Parallel and Distributed Processing Symposium, p. 10 (2006),ISBN: 1-4244-0054-6Google Scholar
  6. 6.
    Juurlink Ben, H.H., Wijshoff Harry, A.G.: A quantitative comparison of parallel computation models. In: Proc. 8th ACM Symp. on Parallel Algorithms and Architectures, pp. 13–24. ACM Press, New York (1996)Google Scholar
  7. 7.
    Chen, W-Y., Bonachea, D., Duell, J., Husbands, P., Iancu, C., Yelick, K.: A Performance Analysis of the Berkley UPC Compiler. In: Annual International Conference on Supercomputing (ICS) (2003)Google Scholar
  8. 8.
    Culler, D., Karp, R., Patterson, D., Sahay, A., Schauser, K.E., Santos, E., Subramonian, R., von Eicken, T.: LogP: Towards a Realistic Model of Parallel Computation. In: PPOPP 1993: ACM SIGPLAN, pp. 1–12. ACM Press, New York (1993)Google Scholar
  9. 9.
    Gerbessiotis, A., Valiant, L.: Direct Bulk-Synchronous Parallel Algorithms. J. of Parallel and Distributed Computing 22, 251–267 (1994)CrossRefGoogle Scholar
  10. 10.
    Gibbons, P.B., Mattias, Y., Ramachandran, V.: Can a Shared-Memory Model Serve as a Bridging Model for Parallel Computation? In: SPAA 1997. 9th Annual ACM Symposium on Parallel Algorithms and Architectures, pp. 72–83. ACM Press, New York (1997)CrossRefGoogle Scholar
  11. 11.
    Valiant, L.G.: A Bridging Model for Parallel Computation. Comm. of the ACM 33(8), 103–111 (1990)CrossRefGoogle Scholar
  12. 12.
    Maggs, B.M, Matheson, L.R., Tarjan, R.E.: Models of Parallel Computation: A Survey and Synthesis. In: Proceeding of the Twenty-Eight Hawaii Conference on System Sciences, vol. 2, pp. 61–70 (1995)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Mohamed Bakhouya
    • 1
  • Jaafar Gaber
    • 2
  • Tarek El-Ghazawi
    • 1
  1. 1.Department of Electrical and Computer Engineering High Performance Computing Laboratory, The George Washington University 
  2. 2.Universite de Technologies de Belfort-Montbeliard 

Personalised recommendations