Extended Overhead Analysis for OpenMP Performance Tuning
Overhead analysis was developed as a performance tuning approach for parallel programming and were adopted by several performance analysis systems for OpenMP programs. In this paper, an extended overhead analysis scheme based on layered model is proposed for OpenMP programming, to further enhance the capability of overhead analysis and thus make the OpenMP performance tuning easier. An example case called ILP/TLP overlap is studied in detail to show the idea of layered overhead model, and a new way to organize the overhead hierarchically is also presented based on the layered overhead model.
KeywordsLayered Model Performance Tuning Chunk Size Abstract Layer Thread Level Parallelism
Unable to display preview. Download preview PDF.
- 1.Performance monitor for IA64. http://www.hpl.hp.com/research/linux/perfmon/
- 2.Intel Inc.: Intel Itanium 2 Processor Reference Manual for Software Development and Optimization (June 2002)Google Scholar
- 3.Open Research Compiler. http://ipf-orc.sourceforge.net.
- 4.G. D. Riley, J. M. Bull, J. R. Gurd: Performance Improvement Through Overhead Analysis: A Case Study in Molecular Dynamics. In Proc. of 11t h Supercomputing (July 1997) 36–43Google Scholar
- 5.J. M. Bull: Measuring Synchronisation and Scheduling Overheads in OpenMP. In Proc. of First European Workshop on OpenMP(EWOMP1999) (September 1999) 99–105.Google Scholar
- 6.J. M. Bull: A hierarchical classification of overheads in parallel programs. In Proc. of First IFIP TC10 International Workshop on Software Engineering for Parallel and Distributed Systems (March 1996) 208–219Google Scholar
- 7.M. K. Bane, G. D. Riley: Automatic Overheads Profiler for OpenMP Codes. In Proc. of the Second European Workshop on OpenMP (EWOMP2000) (September 2000)Google Scholar
- 8.Hong-Linh Truong, Thomas Fahringer: SCALEA: A Performance Analysis Tool for Distributed and Parallel Programs. In Proc. of the 8t h International EuroPar Conf. (August 2002)Google Scholar
- 9.M. K. Bane: Extended Overhead Analysis for OpenMP. In Proc. of the 8t h International EuroPar Conf. (August 2002)Google Scholar
- 10.M. E. Crovella, T. J. LeBlanc: Parallel Performance Prediction Using Lost Cycles Analysis. In Proc. of 8t h Supercomputing (1994) 600–610Google Scholar
- 11.J. K. Hollingsworth: Finding Bottlenecks in Large Scale Parallel Programs. Doctor Thesis, Department of Computer Science, University ofWisconsin-Madison. (1994)Google Scholar
- 12.R. B. Irvin: Mechanisms for Mapping High-Level Parallel Performance Data. In Proc. of the ICPP Workshop on Challenges for Parallel Processing. (August 1996)Google Scholar
- 14.Matthias Müller: Some simple OpenMP optimization techniques. In the Workshop on OpenMP Applications and Tools (WOMPAT 2001) (July 2001)Google Scholar