ScienceDirect® Home Skip Main Navigation Links
You have guest access to ScienceDirect. Find out more.
 
Home
Browse
My Settings
Alerts
Help
 Quick Search
 Search tips (Opens new window)
    Clear all fields    
advertisementadvertisement
Microprocessors and Microsystems
Volume 29, Issue 4, 6 May 2005, Pages 145-153
 
Font Size: Decrease Font Size  Increase Font Size
 Abstract - selected
Article
Purchase PDF (471 K)

  E-mail Article   
  Add to my Quick Links   
Bookmark and share in 2collab (opens in new window)
Request permission to reuse this article
  Cited By in Scopus (0)
 
 
 
Related Articles in ScienceDirect
View More Related Articles
 
View Record in Scopus
 
doi:10.1016/j.micpro.2004.08.001    How to Cite or Link Using DOI (Opens New Window)
Published by Elsevier B.V.

Memory reference caching for activity reduction on address buses

Tony GivargisCorresponding Author Contact Information, E-mail The Corresponding Author and David EppsteinE-mail The Corresponding Author

Department of Computer Science, Center for Embedded Computer Systems, University of California, Irvine, CA 92697-3425, USA

Accepted 5 August 2004. 
Available online 7 September 2004.

Purchase the full-text article



References and further reading may be available for this article. To view references and further reading you must purchase this article.

Abstract

Switching activity on I/O pins of a chip is a measurable contributor to the total energy consumption of the chip. In this work, we present an encoding mechanism that reduces switching activity of external address buses by combining a memory reference caching mechanism with Unit Distance Redundant Codes (UDRC). UDRC are codes that guarantee a Hamming distance of at most one between any pair of encoded symbols. Memory reference caching exploits the fact that memory references are likely to be made up of an interleaved set of sequential streams. Memory reference caching isolates these, otherwise interleaved, streams and limits the communication to an UDRC encoded message that identifies the particular stream, at the cost of at most a single bit-transition. Experiments with 18 embedded system as well as general applications show an average of 58% reduction in switching activity, with the best and worse cases being 86 and 36%, respectively. The maximum performance penalty (i.e. critical-path delay) for a proposed encoder and decoder is 16 and 14 gates, respectively. The area overhead for a proposed encoder and decoder is equivalent to 2033 and 1858 2-input NAND gates, respectively.

Keywords: Bus encoding; Embedded systems; Low power design; Redundant codes

Article Outline

1. Introduction
2. Previous work
3. Approach overview
3.1. System architecture
3.2. Reference caching
3.3. Unit distance redundant codes
3.4. Hardware architecture
4. Experiments
5. Conclusions
References







 
Home
Browse
My Settings
Alerts
Help
Elsevier.com (Opens new window)
About ScienceDirect  |  Contact Us  |  Information for Advertisers  |  Terms & Conditions  |  Privacy Policy
Copyright © 2008 Elsevier B.V. All rights reserved. ScienceDirect® is a registered trademark of Elsevier B.V.