Predicting computation time for advanced processor architectures
- 7 November 2002
- proceedings article
- Published by Institute of Electrical and Electronics Engineers (IEEE)
Abstract
Estimating computation times using analysis techniques is always safe but is becoming prohibitively complex or pessimistic with modern processors. The only alternative approach is to use measurement, but this has the significant disadvantage of optimism - the largest value seen during testing may not be the largest experienced during deployment. In this paper we subject data obtained from measurement to statistical analysis using the techniques of extreme value estimation. A simple case study is described and the approach is illustrated via this study which focuses on the superscalar technique of branch prediction. The approach is applicable to all forms of hardware-induced temporal variability.Keywords
This publication has 9 references indexed in Scilit:
- Bounding completion times of jobs with arbitrary release times and variable execution timesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Computing Maximum Task Execution Times — A Graph-Based ApproachReal-Time Systems, 1997
- Pipelined processors and worst case execution timesReal-Time Systems, 1993
- Predicting program execution times by analyzing static and dynamic program pathsReal-Time Systems, 1993
- Misconceptions about real-time computing: a serious problem for next-generation systemsComputer, 1988
- Residual Life Time at Great AgeThe Annals of Probability, 1974
- Scheduling Algorithms for Multiprogramming in a Hard-Real-Time EnvironmentJournal of the ACM, 1973
- The Bias in Certain Estimates of the Parameters of the Extreme-Value DistributionThe Annals of Mathematical Statistics, 1956
- Limiting forms of the frequency distribution of the largest or smallest member of a sampleMathematical Proceedings of the Cambridge Philosophical Society, 1928