Voltage-clock scaling for low energy consumption in real-Time embedded systems

Yann Hang Lee, C. M. Krishna

Research output: Chapter in Book/Report/Conference proceedingConference contribution

44 Scopus citations

Abstract

Low power and energy consumption will always be an essential requirement in many real-Time embedded applications. Voltage scaling is a relatively novel approach to reducing energy consumption. The idea is that a processor can be run either at high or at low voltage: At high voltage, the clock rate is high but so is the power consumption; at low voltage the clock rate is lower, but the power consumption drops by a greater factor. This immediately suggests a powerful approach to lowering energy consumption in real-Time systems. In this paper we introduce static and dynamic algorithms to control processor voltage to reduce energy consumption. We demonstrate by simulation that our algorithms can significantly reduce total energy consumption.

Original languageEnglish (US)
Title of host publicationProceedings - 6th International Conference on Real-Time Computing Systems and Applications, RTCSA 1999
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages272-279
Number of pages8
ISBN (Electronic)0769503063, 9780769503066
DOIs
StatePublished - 1999
Externally publishedYes
Event6th International Conference on Real-Time Computing Systems and Applications, RTCSA 1999 - Hong Kong, China
Duration: Dec 13 1999Dec 15 1999

Publication series

NameProceedings - 6th International Conference on Real-Time Computing Systems and Applications, RTCSA 1999

Other

Other6th International Conference on Real-Time Computing Systems and Applications, RTCSA 1999
Country/TerritoryChina
CityHong Kong
Period12/13/9912/15/99

ASJC Scopus subject areas

  • Control and Optimization
  • Computer Science Applications
  • Hardware and Architecture
  • Information Systems and Management
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Voltage-clock scaling for low energy consumption in real-Time embedded systems'. Together they form a unique fingerprint.

Cite this