Application execution steering using on-the-fly performance prediction
The execution of an application on a high performance system requires parameters concerning the problem in hand, and those that determine the system mapping, to be specified by a user. The system parameters are typically used to minimise the execution time. However, by the coupling of a performance model with an application, system parameters can be determined without user intervention. In the work presented here, a novel performance prediction system has been used to provide suitable performance models which can determine application mapping parameters, code execution decisions, and system choices on-the-fly. An example compact application of a convolution is used to illustrate the approach for automatically choosing the actual code to be executed, and the number of workstations in a cluster to be utilised. The performance prediction system is shown to have a reasonable accuracy (approximately 10% error), with a rapid evaluation time (typically < 2s).
KeywordsHigh Performance Computing Performance Prediction On-the-fly decision making application steerin
Unable to display preview. Download preview PDF.
- 1.D.A. Agard, Y. Hiraoka, P. Shaw, J.W. Sedat, Fluorescence Micrroscopy in Three Dimensions, in Fluorescence Microscopy of Living Cells in Culture, Elsevier, pp. 353–377, 1989Google Scholar
- 2.J.N.C. Arabe, A.B.B. Lowekamp, E. Saligman, M.Starkey, and P. Stephan, Dome Parallel programming environment in a heterogeneous multi-user environment, Supercomputing, 1995.Google Scholar
- 3.J. Gehring, A. Reinefeld, MARS-A framework for minizing the job execution time in a metacomputing environment, Future Generation Computer Systems, vol. 12, pp. 87–99, 1996Google Scholar
- 4.A. Geist, A. Beguelin, J. Dongarra, W. Jiang, R. Manchek, V. Sunderam, PVM-Parallel Virtual Machine, MIT Press., 1994Google Scholar
- 5.M.W. Hall, J.M. Anderson, et.al, Maximizing Multiprocessor Performance with the SUIF Compiler, IEEE Computer, Vol. 29 (12), December 1996Google Scholar
- 6.D.J. Kerbyson, E. Papaefstathiou, J.S. Harper, S.C. Perry, G.R. Nudd, Is Predictive Tracing Too Late for HPC Users?, in High Performance Computing, R.J. Allan, A. Simpson, D. A. Nicole (Eds), Plenum Press, 1998.Google Scholar
- 7.E. Papaefstathiou, D.J. Kerbyson, G.R. Nudd, T.J. Atherton, An overview of the CHIP3S Performance Prediction Toolset for Parallel Systems, in Proc of 8th ISCA Int. Conf. on Parallel and Distributed Computing Systems, pp. 527–533, Orlando, 1995.Google Scholar
- 8.E. Papaefstathiou, D. J. Kerbyson, G.R. Nudd, A Layered approach to Parallel Software Performance Prediction: A Case Study, in: L. Dekker, W. Smit, and J.C. Zuidervaart, eds., Massively Parallel Processing Applications & Development, pp. 617–624, North-Holland, 1994.Google Scholar
- 9.B. Quin, H.A. Scholl, R.A. Ammar, Micro Time Cosrt Analysis of Parallel Computation, IEEE Trans. on Computers, Vol. 40 (5), pp. 613–628, 1991.Google Scholar
- 10.R. Wolski, Dynamically Forecasting Network Performance Using the Network Weather Service, UCSD Technical Report, TR-CS96-494, 1996.Google Scholar