CS 410/510 Measuring Computer Performance with Dynamic Instrumentation

Spring 2007

Monday/Wednesday 12 noon - 1:50pm
Location: Cramer (CH) 258
Professor Karen Karavanic

Course Note

We will cover the fundamentals of computer performance measurement at a level suitable for advanced undergraduates and beginning graduate students. Topics include measurement tools and techniques, measurement error, benchmarks, and experiment design.

We will focus in particular on tools and technologies related to dynamic instrumentation. Dynamic instrumentation is a technique for inserting instructions into code as it runs (see www.dyninst.org). Course projects will focus on using dynamic instrumentation in a variety of performance measurement contexts: sequential applications, parallel applications, and operating system kernels.

We will have at least one guest lecture, probably two, from folks in the trenches of performance measurement. Homework readings will include textbook chapters and technical papers. Programming projects can be done in pairs or small groups.

Textbook

Required Text: "Measuring Computer Performance: A Practitioner's Guide" by David J. Lilja. Cambridge University Press. ISBN: 0 521 64105 5. This book is now available as an ebook directly from the publishers site: www.cambridge.org. I've also ordered some copies through the PSU bookstore.

Prerequisites (same for graduate students and undergraduate students )

A solid understanding of operating systems and compilers, for example by passing CS321 and CS333 with a grade of B or better. Ability to write, debug, modify, and understand code written in C and C++. Ability to work in a Unix/Linux environment.

Course Mailing List

community_mcp@cs.pdx.edu

Course description
Lecture Schedule
Textbook errors
Exam Information
Grades