Abstract
This paper presents a practical methodology for performance estimation of parallel applications on various supercomputers. Instead of measuring the execution time of the applications with different number of processors on each machine, we estimate the time based on characterization of the workloads. Benchmarking computation and communication primitives in the applications is also performed on parallel computer systems for the characterization. Finally, a performance model is constructed for performance estimation, and verified with different number of processors on each target system.
Our results show that accurate performance estimation is possible with the model constructed using the workload characterization method. With the result figures, we discuss the reasons why over or under estimations occur on each target machine.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Cmelik, B., Keppel, D.: Shade: A Fast Instruction-Set Simulator for Execution Profiling. In: 1994 Sigmetrics, pp. 128–137 (1994)
Davis, H., Goldschmidt, S., Hennessy, J.: Multiprocessor Simulation and Tracing Using Tango. In: Proc. 1991 Int’l Conf. on Parallel Processing, vol. II, pp. 99–107 (1991)
Dincer, K., Bozkus, Z., Ranka, S., Fox, G.: Benchmarking the Computation and Communication Performance of the CM-5. Concurrency: Practice and Experience 8(1), 47–69 (1996)
Giorgi, R., Prete, C.A., Prina, G., Ricciardi, L.: Trace Factory Generating Workloads for Trace-Driven Simulation of Shared-Bus Multiprocessors. IEEE Concurrency 5(4), 54–68 (1997)
http://www.wes.hpc.mil/documentation/ibmspdoc/ibmsp_faqs.html
Kumar, M., So, K.: Trace Driven Simulation for Studying MIMD Parallel Computers. In: Proc. 1989 Int’l Conf. on Parallel Processing, vol. I, pp. 68–72 (1989)
Lauden, J., Lenoski, D.: The SGI Origin: A ccNUMA Highly Scalable Server. In: 24th Int’l Symp. on Computer Architecture, pp. 241–251 (1997)
Liu, L., Peir, J.-K.: A Performance Evaluation Methodology for Coupled Multiple Supercomputers. In: Proc. 1990 Int’l Conf. on Parallel Processing, vol. I, pp. 198–202 (1990)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1999 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
In, J., Jin, C., Peir, J., Ranka, S., Sahni, S. (1999). A Framework for Matching Applications with Parallel Machines. In: Banerjee, P., Prasanna, V.K., Sinha, B.P. (eds) High Performance Computing – HiPC’99. HiPC 1999. Lecture Notes in Computer Science, vol 1745. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-46642-0_48
Download citation
DOI: https://doi.org/10.1007/978-3-540-46642-0_48
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-66907-4
Online ISBN: 978-3-540-46642-0
eBook Packages: Springer Book Archive