Factor Analysis of Applications Performance Data for the Cray Y-MP

Abstract

Factor analysis was applied to supercomputer applica tions performance data recorded by the hpm hardware performance monitor on the Cray Y-MP for a sample of 146 user codes. The analysis reduced the volumi nous hpm data to five dominant factors which were used as performance metrics for applications code ex ecution on the Y-MP. Case studies of Fortran kernels demonstrated that the proposed metrics accurately reflect changes in performance due to source code modifications. Therefore, the factor-analysis model so lution provides a supercomputer performance metric that is both parsimonious and simple to interpret as an alternative to the complexities of hpm-derived perfor mance attributes. Metrics such as these should prove to be of value in code optimization, user training, and characterization of site-specific workloads.

Get full access to this article

View all access options for this article.

References

Berry, M. , Chen, D. , Koss, P. , et al. 1989. The perfect club benchmarks: effective performance evaluation of supercomputers. Internat. J. Supercomp. App . 3(3):5-40.

Bradley, D. , Cybenko, G. , Gao, H. , Ahmad, F. , Golab, J. , Larson, J. , and Straka, M. 1991. Supercomputer workload decomposition and analysis. In Proceedings of the A CM international conference on supercomputing. New York: ACM Press, pp. 458-467.

Bradley, D.K. , and Larson, J.L. 1991. Fine-grained measurement of loop performance on the Cray Y-MP. In Fifth SIAM conference on parallel processing for scientific applications, edited by Danny C. Sorenson Philadelphia, Pennsylvania : SIAM.

Cray Research Inc. 1989. UNICOS performance utilities reference manual SR-2040, version B (UNICOS 5.0). Mendota Heights, Minnesota.

Cray Research Inc. 1991. UNICOS performance utilities reference manual SR-2040, version 6.0. Mendota Heights, Minnesota.

Cray Research Inc. 1993. CRAY C90 series functional description manual HR-04028-0A. Mendota Heights, Minnesota.

Delic, G. 1993. Performance attributes for code and workload analysis on Cray X-MP and Y-MP systems. Internat. J. of Supercomp. App. 7(4): 304-336.

Harman, H.H. 1976. Modern factor analysis, 3rd ed. Chicago: University of Chicago Press .

Infante, E.F. 1986. An agenda for improved evaluation of supercomputer performance, A report prepared by the committee on supercomputer performance and development . Washington, D.C.: National Academy Press.

10.

Kim, J.-O. , and Mueller, C.W. 1978a. Factor analysis: statistical methods and practical issues . Beverly Hills, California: Sage Publications Ltd.

11.

Kim, J.-O. , and Mueller, C.W. 1978b. Introduction to factor analysis: what it is and how to do it. Beverly Hills, California: Sage Publications Ltd.

12.

Levesque, J.M. , and Williamson, J.W. 1989. A guidebook to Fortran on supercomputers. New York: Academic Press.

13.

Mulaik, S.A. 1972. The foundations of factor analysis. New York: McGraw-Hill.

14.

National Science Foundation, Division of Advanced Scientific Computing. 1989. Report of the Workshop on Scientific Computing Performance Analysis, August 29-31, Boulder, Colorado (unpublished).

15.

Reyment, R. , and jöreskog, K.G. 1993. Applied factor analysis in the natural sciences . New York: Cambridge University Press.

16.

Saavedra-Barrera, R.H. , Smith, A.J. , and Miya, E. 1989. Machine characterization based on an abstract high-level language machine. IEEE Trans. Gomp. 38(12):1659-1678.

17.

SAS Institute Inc. 1989. SAS/STAT user's guide, version 6, fourth edition, volume 1. Cary, North Carolina.

18.

Zima, H. , and Chapman, B. 1991. Supercompilers for vector and parallel computers . New York: ACM Press.