Metrics of Computer Performance
David J. Lilja, Ph.D., P.E.
Before you can
begin to determine anything about the performance of a computer system, you
must decide what aspects of the system are interesting and useful to measure.
In this course, you will learn about the common metrics that are used to quantify
a system's performance, the characteristics of a good metric, the meaning of
speedup and relative change, and how to measure the fundamental metric of performance
This course includes
a multiple-choice quiz at the end, which is designed to enhance the understanding
of the course materials.
After completing this 2-hour course, you will be able to:
The reading assignment for this course is Chapter 2 of Measuring Computer Performance: A Practitioner's Guide, David J. Lilja, Cambridge University Press, 2000.
If you don't have this book, you can purchase Chapter 1 in PDF format online at eBooks.com for a modest cost. The price for this course listed on this website does not include the cost of purchasing the chapter through eBooks.com. However, the price has been reduced to compensate for the cost of purchasing the chapter required. If you plan to take all 6 courses (E132 to E137) based on this book, you may consider to purchase a hard copy of the book or the entire book in PDF format online through eBooks.com.
We begin our study of computer systems performance analysis by examining those things that are useful to measure -- that is, the metrics of performance. There are essentially three types of characteristics of a computer system that we can measure directly:
1. A count of how
many times an event occurs.
2. The duration of some time interval.
3. The size of some parameter.
Usually, we are interested in how quickly a given program is executed. As a result, the fundamental performance metric of any system is time. We can compare times directly, or we can use them to derive appropriate rates. But without a precise and accurate measure of time, it is impossible to analyze or compare any system performance characteristics. The basic technique for measuring time in a computer system is analogous to using a stopwatch to measure the time required to perform some event. Some examples of common interval timers are provided here.
While time is usually
considered the ultimate measure of performance, there are many other performance
metrics that are in common use. Not all of these metrics are good, however,
in that they sometimes can lead to erroneous conclusions. After completing this
course, you should understand what makes a good performance metric. Furthermore,
you should have developed some sense for why it is so difficult to develop good
performance metrics, and why, therefore, it is so difficult to accurately and
rigorously compare the performance of computer systems.
Once you finish studying the above course content, you need to take a quiz to obtain the PDH credits.