文献类型：期刊文献

英文题名：An energy optimization method for time-constrained applications on multi-core systems

作者：Wang, Yingfeng[1,2]; Liu, Zhijing[2]; Yao, Yong[2]

第一作者：王颖锋;Wang, Yingfeng

通讯作者：Wang, Y.

机构：[1] College of Computer and Information Engineering, Henan University Of Economics And Law, Zhengzhou 450002, China; [2] School of Computer Science and Technology, Xidian University, Xi'an, 710071, China

第一机构：河南财经政法大学计算机与信息工程学院

年份：2011

卷号：6

期号：12

起止页码：238-247

外文期刊名：Journal of Convergence Information Technology

收录：EI(收录号：20115214650638);Scopus(收录号：2-s2.0-84455202868)

语种：英文

外文关键词：Constrained optimization - Embedded systems - Energy conservation - Energy utilization - Green computing - Iterative methods - Scheduling - Voltage scaling

摘要：In this paper, a new scheduling approach to minimize energy consumption of multi-core systems is proposed considering transition overhead between performance modes and communication overhead between processor cores. The proposed approach is designed to achieve energy reduction based on dynamic voltage scaling and adaptive body bias for applications which can be modeled as directed acyclic graphs with hard timing constraints. It takes the corresponding retimed directed acyclic graphs of original directed acyclic graphs as one of input parameters. First, the proposed method determines initial task order and frequency assignment to get the minimum initial scheduling length under given timing constraint. Then it iteratively selects candidate task and scales the frequency of the candidate task to achieve the maximum ratio of reduced energy to increased execution time. In addition, it regulates task order so that more slacks can be generated to reduce energy consumption. Simulations have been carried out with several task sets based on the power model of 70nm technology. The results demonstrate that our approach is effective in energy optimization compared with previous work.