Measuring Computer Performance: A Practitioner's Guide

by David J. Lilja

• ISBN13: 9780521641050
• ISBN10: 0521641055
• Format: Hardcover
• Copyright: 2000-08-15
• Publisher: Cambridge University Press

Summary

Measuring Computer Performance sets out the fundamental techniques used in analyzing and understanding the performance of computer systems. The emphasis is on practical methods of measurement, simulation, and analytical modeling. David Lilja discusses performance metrics and provides detailed coverage of the strategies used in benchmark programs. In addition to intuitive explanations of key statistical tools, he describes the general "design of experiments" technique and shows how the maximum amount of information can be obtained with minimum effort. Features include appendices listing common probability distributions and statistical tables and a glossary of important technical terms.

Author Biography

David J. Lilja is a Professor in the Department of Electrical and Computer Engineering at the University of Minnesota in Minneapolls.

Table of Contents

Preface	xi
Acknowledgements	xv
Introduction	1	(8)
Measuring performance	1	(1)
Common goals of performance analysis	2	(2)
Solution techniques	4	(3)
Summary	7	(1)
Exercises	7	(2)
Metrics of performance	9	(16)
What is a performance metric?	9	(1)
Characteristics of a good performance metric	10	(2)
Processor and system performance metrics	12	(7)
Other types of performance metrics	19	(1)
Speedup and relative change	19	(2)
Means versus ends metrics	21	(2)
Summary	23	(1)
For further reading	23	(1)
Exercises	24	(1)
Average performance and variability	25	(18)
Why mean values?	25	(1)
Indices of central tendency	26	(3)
Other types of means	29	(6)
Quantifying variability	35	(4)
Summary	39	(1)
For further reading	40	(1)
Exercises	41	(2)
Errors in experimental measurements	43	(18)
Accuracy, precision, and resolution	43	(2)
Sources of errors	45	(1)
A model of errors	46	(1)
Quantifying errors	47	(10)
Summary	57	(1)
For further reading	57	(1)
Exercises	58	(3)
Comparing alternatives	61	(21)
Comparing two alternatives	62	(9)
Comparing more than two alternatives	71	(9)
Summary	80	(1)
For further reading	80	(1)
Exercises	81	(1)
Measurement tools and techniques	82	(29)
Events and measurement strategies	82	(4)
Interval timers	86	(6)
Program profiling	92	(5)
Event tracing	97	(7)
Indirect and ad hoc measurements	104	(1)
Perturbations due to measuring	105	(2)
Summary	107	(1)
For further reading	108	(1)
Exercises	109	(2)
Benchmark programs	111	(28)
Types of benchmark programs	112	(5)
Benchmark strategies	117	(8)
Example benchmark programs	125	(7)
Summary	132	(3)
For further reading	135	(2)
Exercises	137	(2)
Linear-regression models	139	(18)
Least-squares minimization	139	(3)
Confidence intervals for regression parameters	142	(3)
Correlation	145	(3)
Multiple linear regression	148	(2)
Verifying linearity	150	(1)
Nonlinear models	151	(4)
Summary	155	(1)
For further reading	155	(1)
Exercises	156	(1)
The design of experiments	157	(24)
Types of experiments	158	(1)
Terminology	158	(1)
Two-factor experiments	159	(9)
Generalized m-factor experiments	168	(4)
n2m experiments	172	(5)
Summary	177	(1)
For further reading	177	(1)
Exercises	178	(3)
Simulation and random-number generation	181	(36)
Simulation-efficiency considerations	182	(1)
Types of simulations	183	(7)
Random-number generation	190	(13)
Verification and validation of simulations	203	(9)
Summary	212	(1)
For further reading	213	(1)
Exercises	214	(3)
Queueing analysis	217	(22)
Queueing-network models	218	(2)
Basic assumptions and notation	220	(1)
Operational analysis	221	(4)
Stochastic analysis	225	(10)
Summary	235	(2)
For further reading	237	(1)
Exercises	237	(2)
Appendix A Glossary	239	(3)
Appendix B Some useful probability distributions	242	(7)
Appendix C Selected statistical tables	249	(9)
Index	258