Skip to main content
Log in

Workload characterization using the CRAY hardware performance monitor

  • Published:
The Journal of Supercomputing Aims and scope Submit manuscript

Abstract

This paper describes some preliminary results about a workload characterization study at a national supercomputer center. This study is part of a larger project to develop an analytic methodology for the construction of arepresentative benchmark set. Our study reports on 292,254 user jobs during the period from June 1992 to July 1993 and represents about one-half of the wall clock time available on the four-processor CRAY Y-MP at the National Center for Supercomputing Applications during this time. The performance statistics we gathered in some sensedefine the workload because of the large fraction of the available time that was recorded. These statistics will allow us to “reverse engineer” a set of benchmarks that analytically represents the workload at this site.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

References

  • Berry, M.W. 1991. Scientific workload characterization by loop-based analyses. Tech. Rept. CS-91-140, Univ. of Tenn., Knoxville, Tenn.

    Google Scholar 

  • Berry, M., Cybenko, G., and Larson, J. 1991. Scientific benchmarks characterization.Parallel Computing, 17, 10 & 11 (Dec.): 1173–1194.

    Google Scholar 

  • Berry, M., et al. 1989. The PERFECT Club benchmarks: Effective performance evaluation of supercomputers.Internat. J. Supercomputer Applications, 3, 3: 5–40.

    Google Scholar 

  • Bradley, D.K., and Larson, J.L. 1993. A parallelism-based analytic approach to performance evaluation using application programs.Proc. IEEE, 81, 8 (Aug.): 1126–1135.

    Google Scholar 

  • Bradley, D., Cybenko, G., Gao, H., Larson, J., Ahmad, F., Golab, J., and Straka, M. 1991. Supercomputer workload decomposition and analysis. InProc., ACM Internat. Conf. on Supercomputing, pp. 458–467.

  • Bucher, I., and Martin, J. 1982. Methodology for characterizing a scientific workload. Tech. Rept. LA-UR-82-1702, Los Alamos Nat. Laboratory, Los Alamos, N. Mex.

    Google Scholar 

  • Ciment, M., Scherlis, W., Griffin, S., Browstein, C., et al., eds. 1993.Grand Challenges 1993: High Performance Computing and Communications. Nat. Sci. Fndtn.

  • Cray Research. 1991.UNICOS Performance Utilities Manual. SR-2040, 6.0 ed.

  • Delic, G. 1991. Performance analysis of a 24 code sample on Cray X/Y-MP systems at the Ohio Supercomputer Center. InProc., 5th SIAM Conf. on Parallel Processing for Scientific Applications, pp. 530–535.

  • Dongarra, J. J., Meuer, H.W., and Strohmaier, E., eds. 1994.TOP500 Report: 1993. Univ. of Mannheim, Germany.

    Google Scholar 

  • Ferrari, D. 1972. Workload characterization and selection in computer performance measurement.Computer, 5, 4 (Aug): 18–24.

    Google Scholar 

  • Gaede, S.L. 1981. Tools for research in computer workload characterization and modeling. InExperimental Computer Performance and Evaluation (D. Ferrari and M. Spadoni, eds.), North-Holland, pp. 235–247.

  • Grenander, U., and Tsao, R.F. 1971. Quantitative methods for evaluating computer system performance: A review and proposals. InProc., Conf. on Statistical Methods for the Evaluation of Computer Systems Performance (Brown Univ.).

  • Larson, J.L. 1991. Collecting and interpreting HPM performance data on the CRAY Y-MP.NCSA Datalink, 5, 5 (Nov.–Dec.): 14–24.

    Google Scholar 

  • Larson, J., and Lutz, B. 1985.PERFTRACE User's Guide. Tech Rept., Cray Research, Inc.

  • Marcusiu, D. 1991. Performance data now automatic on CRAY Y-MP.NCSA Datalink, 5, 4 (Sept.–Oct.): 5–6.

    Google Scholar 

  • Martin, J., Bucher, I., and Warnock, T. 1983. Workload characterization for vector computers: Tools and techniques. Tech. Rept. LA-UR-83-305, Los Alamos Nat. Laboratory, Los Alamos, N. Mex.

    Google Scholar 

  • Nelson, H. 1985. Using the performance monitors on the X-MP/48.LLNL Tentacle.

  • Rosen, S. 1976.Lectures on the Measurement and Evaluation of the Performance of Computing Systems. Regional Conf. Series in Applied Mathematics, SIAM.

  • Sato, D. 1988. SCD conducts performance and benchmarking tests.SCD Computing News, 9, 4 (Apr.): 3–5.

    Google Scholar 

  • Scientific Supercomputing Subcommittee. 1992. NSF supercomputer centers study. Tech. Rept., IEEE Comp. Soc. (Feb.).

  • West, E. 1989. X-MP system wide performance analysis.The CLSC News, 3, 1 (Apr.): 15–17.

    Google Scholar 

  • Williams, E.A. 1988. Measurement of two scientific workloads using the CRAY X-MP performance monitor. Tech. Rept. SRC-TR-88-020, Center for Computing Sciences, Bowie, Md.

    Google Scholar 

  • Williams, E., Myers, C.T., and Koskela, R. 1990. The characterization of two scientific wordloads using the CRAY X-MP hardware performance monitor. InProc., Supercomputing '90, pp. 142–152.

Download references

Author information

Authors and Affiliations

Authors

Rights and permissions

Reprints and permissions

About this article

Cite this article

Gao, H., Larson, J.L. Workload characterization using the CRAY hardware performance monitor. J Supercomput 9, 391–412 (1995). https://doi.org/10.1007/BF01206274

Download citation

  • Received:

  • Accepted:

  • Issue Date:

  • DOI: https://doi.org/10.1007/BF01206274

Keywords

Navigation