By Topic

Techniques for Real-System Characterization of Java Virtual Machine Energy and Power Behavior

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Gilberto Contreras ; Department of Electrical Engineering, Princeton University. ; Margaret Martonosi

The Java platform has been adopted in a wide variety of systems ranging from portable embedded devices to high-end commercial servers. As energy, power dissipation, and thermal challenges begin to affect all design spaces, Java Virtual Machines will need to evolve in order to respond to these and other emerging issues. Developing a power-conscious Java runtime system begins with a detailed per-component understanding of the energy, performance and power behavior of the system, as well as each component's impact on overall application execution. This paper presents techniques for characterizing Java power and performance, as well as results from applying these techniques to the Jikes RVM, for some of the most salient Java Virtual Machine components. Components studied include the garbage collector, the class loader, and the runtime compilation subsystem. Real-system measurements with our efficient, low-perturbation infrastructure offer valuable insights that can aid virtual machine designers in improving energy-efficiency. For example, our results show that JVM energy consumption can comprise as much as 60% of the total energy consumed. In addition, we find that generational garbage collectors offer the best energy-performance for small heap sizes and that this efficiency is challenged by non-generational collectors for large heaps. Overall, given the rising importance of Java systems and of power/thermal challenges, this paper's detailed real-systems examination can lend useful insights for many real-world systems

Published in:

2006 IEEE International Symposium on Workload Characterization

Date of Conference:

25-27 Oct. 2006