A framework for power estimation and reduction in multi-core architectures using basic block approach
by M. Rajasekhara Babu; P. Venkata Krishna; M. Khalid
International Journal of Communication Networks and Distributed Systems (IJCNDS), Vol. 10, No. 1, 2013

Abstract: In recent scenarios, power consumption is critical for battery operated devices. There are wide varieties of implementations of dynamic voltage scaling (DVS) algorithm to reduce energy or power. This paper presents a framework called PERMA, power estimator and reducer for multi-core architectures. The PERMA estimates power consumption and suggests analytical procedure to reduce power consumption at basic block level rather than at region level using clock cycles of instructions for a particular architecture (x86). PERMA estimates execution time for each basic block for various voltage levels and chooses best out of these. Therefore, PERMA evaluates the extent to which the voltage can be varied for various Basic Blocks to reduce power consumption without degrading execution time. Finally, it is tested for matrix multiplication of various sizes. There is an improvement in the execution time up to 33.43% with PERMA and 21.89% without PERMA.

Online publication date: Fri, 28-Jun-2013

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Communication Networks and Distributed Systems (IJCNDS):
Login with your Inderscience username and password:

    Username:        Password:         

Forgotten your password?


Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com