Abstract
The design and evaluation of high performance computers has concentrated on increasing computational speed for applications. This performance is often measured on a well configured dedicated system to show the best case. In the real environment, resources are not always dedicated to a single task, and systems run tasks that may influence each other, so run times vary, sometimes to an unreasonably large extent. This paper explores the amount of variation seen across four large distributed memory systems in a systematic manner. It then analyzes the causes for the variations seen and discusses what can be done to decrease the variation without impacting performance.
Chapter PDF
Similar content being viewed by others
Keywords
- Performance Variation
- Parallel Architecture
- Architectural Feature
- Lawrence Berkeley National Laboratory
- Performance Variability
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
Woo, Stephan, et al. The SPLASH-2 Programs: Characterization and Methological Considerations, Proc ISCA, June 1995, Santa Margherita Ligure, Italy, pp. 24–36.
Henning, John L., SPEC CPU 2000: Measuring CPU performance in the New Millenium, IEEE Computer, July 2000, pp. 28–35.
David H. Bailey, et al., The NAS Parallel Benchmarks, Intl. Journal of Supercomputer Applications, vol. 5, no. 3 (Fall 1991), pp. 66–73 http://www.nas.nasa.gov/Research/Reports/Techreports/1996/nas-96-010-abstract.html
William T.C. Kramer and Clint Ryan, Performance Varibility of Highly Parallel Architectures, National Energy Research Scientific Computing Center, Lawrence Berkeley National Laboratory Report LBNL-52591, March 2003. http://www.nersc.gov/ kramer/variation-systems.pdf
Cray T3E System Support Skills, Cray Research Technical documentation, Number R-T3ESSS, Revision R/H, May 1997
UNICOS/mk Resource Administration, SG-2602, version 2.0.2
IBM Parallel Environment for AIX Dynamic Probe Class Library Programming Guide, Version 3 Release 1, Document Number SA22-7420-00
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Kramer, W.T.C., Ryan, C. (2003). Performance Variability of Highly Parallel Architectures. In: Sloot, P.M.A., Abramson, D., Bogdanov, A.V., Gorbachev, Y.E., Dongarra, J.J., Zomaya, A.Y. (eds) Computational Science — ICCS 2003. ICCS 2003. Lecture Notes in Computer Science, vol 2659. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-44863-2_55
Download citation
DOI: https://doi.org/10.1007/3-540-44863-2_55
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-40196-4
Online ISBN: 978-3-540-44863-1
eBook Packages: Springer Book Archive