Efficient DVFS to prevent hard faults for many-core architectures

1Citations
Citations of this article
5Readers
Mendeley users who have this article in their library.

This article is free to access.

Abstract

Dynamic Voltage and Frequency Scaling (DVFS) is a widely-used and efficient technology for Dynamic Power management (DPM). To avoid hard faults caused by voltage and frequency scaling, some overhead always be imposed on the performance of applications due to the latency of DVFS. Besides, on many-core architectures, the design of multiple voltage domains has made the latency of DVFS a much more significant issue. In this paper, we propose an efficient DVFS scheme to prevent hard faults, meanwhile eliminating the impact of latency of DVFS as possible. The main idea is applying Retroactive Frequency Scaling (RFS) where the latency of DVFS might be introduced. Based on the analysis, our approach is expected to achieve noticeable performance improvement on many-core architectures. © 2014 IFIP International Federation for Information Processing.

Cite

CITATION STYLE

APA

Lai, Z., Zhao, B., & Su, J. (2014). Efficient DVFS to prevent hard faults for many-core architectures. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8407 LNCS, pp. 674–679). Springer Verlag. https://doi.org/10.1007/978-3-642-55032-4_69

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free