Approaches to parallel performance prediction
Fred Howell
PhD Thesis, Dept of Computer Science,
University of Edinburgh, 1996
The chapters of my thesis are available separately below, in
pdf and postscript versions. Unfortunately, the two interesting chapters
are only available in postscript because of problems with
the pdf generator.
Abstract
Designing parallel programs is both interesting and difficult.
The reason for using a parallel machine is to obtain better performance,
but the programmer will have little idea of the performance of
a program at design time, and will only find out by actually
running it. Design decisions have to be be made by guesswork
alone.
This thesis explores an alternative by providing data sheets
describing the performance of parallel building blocks, and then
seeing how they may be used in practice.
The simplest way of using the data sheets is based on a graphing
and equation plotting tool. More detailed design information is
available from a ``reverse'' profiling technique which adapts
standard profiling to generate predictions rather than measurements.
The ultimate method for prediction is based on discrete event
simulation, which allows modelling of all programs but is the most
complex to use.
The methods are compared, and their suitability for different
design problems is discussed.
Contents
- Introduction
(.ps)
- Experimental Approach
(.ps)
- The Performance Characterisation of MPI Functions
(.ps only)
- Simple Performance Estimates of the Cowichan Problems
(.ps)
- Reverse Profiling (.ps only)
- A Simulation Tool for MPI Performance Prediction
(.ps)
- Conclusion
(.ps)
Appendices
- An overview of MPI
(.ps)
- An example MPI datasheet, title and
contents (.ps only)
(see also the datasheet web page for html)
- References of papers and full text of two (copies of others on request):
-
F.W. Howell.
Reverse profiling. In Innes Jelly and Ian Gorton, editors,
Software Engineering
for Parallel and Distributed Systems : Proceedings of the First IFIP TC10
International Workshop on Parallel and Distributed Software Engineering,
pages 244-255. IFIP, Chapman and Hall, March 1996.
(and in .ps)
- F.W. Howell and R.N. Ibbett.
STATE OF THE ART IN PERFORMANCE MODELLING AND SIMULATION
Modelling and Simulation of Advanced Computer Systems: Techniques, Tools and
Tutorials, edited by Kallol Bagchi,
Chapter 1 : Hierarchical Architecture
Simulation Environment, pages 1-18.
Gordon and Breach, 1996. (and in .ps)
Fred Howell
Last modified: Mon Jun 30 14:23:18 BST