Abstract
Accurately modeling and predicting performance for large-scale applications becomes increasingly difficult as system complexity scales dramatically. Analytic predictive models are useful, but are difficult to construct, usually limited in scope, and often fail to capture subtle interactions between architecture and software. In contrast, we employ multilayer neural networks trained on input data from executions on the target platform. This approach is useful for predicting many aspects of performance, and it captures full system complexity. Our models are developed automatically from the training input set, avoiding the difficult and potentially error-prone process required to develop analytic models. This study focuses on the high-performance, parallel application SMG2000, a much studied code whose variations in execution times are still not well understood. Our model predicts performance on two large-scale parallel platforms within 5%-7% error across a large, multi-dimensional parameter space.
Chapter PDF
Similar content being viewed by others
Keywords
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.
References
Brown, P., Falgout, R.D., Jones, J.E.: Semicoarsening multigrid on distributed memory machines. SIAM J. Sci. Computing 21, 1823–1834 (2000)
Carrington, L., Snavely, A., Gao, X., Wolter, N.: A performance prediction framework for scientific applications. In: International Conference on Computational Science Workshop on Performance Modeling and Analysis (PMA 2003), June 2003, pp. 926–935 (2003)
Carrington, L., Wolter, N., Snavely, A., Lee, C.B.: Applying an automatic framework to produce accurate blind performance predictions of full-scale HPC applications. In: Department of Defense Users Group Conference (June 2004)
Falgout, R.D., Yang, U.M.: Hypre: a Library of High Performance Preconditioners. In: Sloot, P.M.A., Tan, C.J.K., Dongarra, J., Hoekstra, A.G. (eds.) ICCS-ComputSci 2002. LNCS, vol. 2331, pp. 632–641. Springer, Heidelberg (2002)
Karkhanis, T.S., Smith, J.E.: A first-order superscalar processor model. In: Proceedings of the 31st Annual International Symposium on Computer Architecture, June 2004, pp. 338–349 (2004)
Kerbyson, D.J., Alme, H.J., Hoisie, A., Petrini, F., Wasserman, A.J., Gittings, M.: Predictive performance and scalability modeling of a large-scale application. In: Proceedings of IEEE/ACM Supercomputing 2001 (November 2001)
Marin, G., Mellor-Crummey, J.: Cross-architecture performance predictions for scientific applications using parameterized models. In: Proceedings of the International Conference on Measurement and Modeling of Computer Systems (Sigmetrics 2004), June 2004, pp. 2–13 (2004)
Mitchell, T.M.: Machine Learning. WCB/McGraw Hill, Boston (1997)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Ipek, E., de Supinski, B.R., Schulz, M., McKee, S.A. (2005). An Approach to Performance Prediction for Parallel Applications. In: Cunha, J.C., Medeiros, P.D. (eds) Euro-Par 2005 Parallel Processing. Euro-Par 2005. Lecture Notes in Computer Science, vol 3648. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11549468_24
Download citation
DOI: https://doi.org/10.1007/11549468_24
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-28700-1
Online ISBN: 978-3-540-31925-2
eBook Packages: Computer ScienceComputer Science (R0)