Predictive temperature-aware DVFS

Jong Sung Lee, Kevin Skadron, Sung Woo Chung

Research output: Contribution to journalArticlepeer-review

82 Citations (Scopus)


In this paper, we propose predictive temperature-aware Dynamic Voltage and Frequency Scaling (DVFS) using the performance counters that are already embedded in commercial microprocessors. By using the performance counters and simple regression analysis, we can predict the localized temperature and efficiently scale the voltage/frequency. When localized thermal problems that were not detected by thermal sensors are found after layout (or fabrication), the thermal problems can be avoided by the proposed software solution without delaying time-to-market. The evaluation results show that in a Linux-based laptop with the Intel Core2 Duo processor, DVFS using the performance counters performs comparable to DVFS using the thermal sensor.

Original languageEnglish
Article number5255230
Pages (from-to)127-133
Number of pages7
JournalIEEE Transactions on Computers
Issue number1
Publication statusPublished - 2010

Bibliographical note

Funding Information:
This paper is an extended version of [3]. This work was supported by the Korea Research Foundation Grant of the Korean Government (KRF-2006-331-D00452), and the Korea Science and Engineering Foundation (KOSEF) grant funded by the Korea government (MEST) (No. R01-2007-000-20750-0). This work was also funded by the US National Science Foundation under grant no. EHS-0509245 and a grant from Intel MRL. The authors would like to thank Professor Gabriel Loh for providing Core 2 Duo related parameters. Finally, they would like to thank the anonymous reviewers for their helpful feedback. Sung Woo Chung is the corresponding author of this paper.


  • DVFS
  • Dynamic thermal management
  • Microprocessor
  • Performance counter
  • Thermal sensor

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Hardware and Architecture
  • Computational Theory and Mathematics


Dive into the research topics of 'Predictive temperature-aware DVFS'. Together they form a unique fingerprint.

Cite this