By Topic

A Near-optimal Solution for the Heterogeneous Multi-processor Single-level Voltage Setup Problem

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

3 Author(s)
Tai-Yi Huang ; National Tsing Hua University, Dept. of Computer Science, Hsinchu, Taiwan 300, R.O.C. ; Yu-Che Tsai ; Edward T. -H. Chu

A heterogeneous multi-processor (HeMP) system consists of several heterogeneous processors, each of which is specially designed to deliver the best energy-saving performance for a particular category of applications. A low-power real-time scheduling algorithm is required to schedule tasks on such a system to minimize its energy consumption and complete all tasks by their deadline. The problem of determining the optimal speed for each processor to minimize the total energy consumption is called the voltage setup problem. This paper provides a near-optimal solution for the HeMP single-level voltage setup problem. To our best knowledge, we are the first work that addresses this problem. Initially, each task is assigned to a processor in a local-optimal manner. We next propose a couple of solutions to reduce energy by migrating tasks between processors. Finally, we determine each processor's speed by its final workload and the deadline. We conducted a series of simulations to evaluate our algorithms. The results show that the local-optimal partition leads to a considerably better energy-saving schedule than a commonly-used homogeneous multi-processor scheduling algorithm. Furthermore, at all measurable configurations, our energy consumption is at most 3% more than the optimal value obtained by an exhaustive iteration of all possible task-to-processor assignments. In summary, our work is shown to provide a near-optimal solution at its polynomial-time complexity.

Published in:

2007 IEEE International Parallel and Distributed Processing Symposium

Date of Conference:

26-30 March 2007