Advertisement

Importance of Runtime Considerations in Performance Engineering of Large-Scale Distributed Graph Algorithms

  • Jesun Sahariar FirozEmail author
  • Thejaka Amila Kanewala
  • Marcin Zalewski
  • Martina Barnas
  • Andrew Lumsdaine
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9523)

Abstract

Due to the ever increasing complexity of the modern supercomputers, performance analysis of irregular applications became an experimental endeavor. We show that runtime considerations are inseparable from algorithmic concerns in performance engineering of large-scale distributed graph algorithms, and we argue that the whole system stack, starting with the algorithm at the top down to low-level communication libraries must be considered.

Keywords

Runtime System Message Type Distribute Control Active Message Breadth First Search 
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 research used Big Red2 (Funded by Lilly Endowment, Inc. and Indiana METACyt Initiative). Support by NSF grant 1111888 gratefully acknowledged.

References

  1. 1.
    http://hpx.crest.iu.edu/. Accessed 25 May 2015
  2. 2.
    Big Red II at Indiana University. http://rt.uits.iu.edu/ci/systems/BRII.php#info. Accessed 17 Apr 2015
  3. 3.
    Kissel, E., Swany, M.: Session layer burst switching for high performance data movement. In: Proceedings of the 8th International Workshop on Protocols for Future, Large-Scale & Diverse Network Transports (2010)Google Scholar
  4. 4.
    Meyer, U., Sanders, P.: \({\varDelta }\)-stepping: a parallelizable shortest path algorithm. J. Algorithms 49(1), 114–152 (2003)zbMATHCrossRefMathSciNetGoogle Scholar
  5. 5.
    Murphy, R.C., Wheeler, K.B., Barrett, B.W., Ang, J.A.: Introducing the graph 500 benchmark. Cray User’s Group (CUG) (2010)Google Scholar
  6. 6.
    Pingali, K., et al.: The Tao of Parallelism in Algorithms. ACM SIGPLAN Not. 46(6), 12–25 (2011)CrossRefGoogle Scholar
  7. 7.
    Willcock, J.J., Hoefler, T., Edmonds, N.G., Lumsdaine, A.: AM++: a generalized active message framework. In: Proceedings of the 19th International Conference on Parallel Architectures and Compilation Techniques, pp. 401–410. ACM (2010)Google Scholar
  8. 8.
    Willcock, J.J., Hoefler, T., Edmonds, N.G., Lumsdaine, A.: Active pebbles: a programming model for highly parallel fine-grained data-driven computations. In: Proceedings of the 16th ACM Symposium on Principles and Practice of Parallel Programming, pp. 305–306. ACM (2011)Google Scholar
  9. 9.
    Zalewski, M., Kanewala, T.A., Firoz, J.S., Lumsdaine, A.: Distributed control: priority scheduling for single source shortest paths without synchronization. In: Proceedings of the Fourth Workshop on Irregular Applications: Architectures and Algorithms, pp. 17–24. IEEE (2014)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Jesun Sahariar Firoz
    • 1
    Email author
  • Thejaka Amila Kanewala
    • 1
  • Marcin Zalewski
    • 1
  • Martina Barnas
    • 1
  • Andrew Lumsdaine
    • 1
  1. 1.Center for Research in Extreme Scale Technologies (CREST)Indiana UniversityBloomingtonUSA

Personalised recommendations