标题:Fast and Accurate Dynamic Power Estimation method based on Component
作者:Li Zhao; Li Yede; Wu Xinghua
通讯作者:Li, Z;Li, YD;Wu, XH
作者机构:[Li Zhao; Li Yede] Shandong Univ Technol, Coll Comp Sci & Technol, Zibo, Peoples R China.; [Wu Xinghua] Shandong Univ Technol, Sch Elect & Elect Eng 更多
会议名称:4th National Conference on Electrical, Electronics and Computer Engineering (NCEECE)
会议日期:DEC 12-13, 2015
来源:PROCEEDINGS OF THE 2015 4TH NATIONAL CONFERENCE ON ELECTRICAL, ELECTRONICS AND COMPUTER ENGINEERING ( NCEECE 2015)
出版年:2016
卷:47
页码:1025-1030
关键词:dynamic power estimation; component; calculation accuracy; calculation; speed
摘要:Dynamic power optimization is an important part of the power optimization for embedded parallel processing. Dynamic power estimation is the premise for dynamic power optimization. The evaluation model with fast calculating speed and high accuracy can improve the efficiency for dynamic power optimization. The dynamic power evaluation methods based on low level simulation have high accuracy. But it is very time consuming. The high level dynamic power estimation models have higher speed, but have lower calculation accuracy. For this issue, the paper proposes a dynamic power estimation method based on component which combines ow level dynamic power evaluation methods with high level dynamic power estimation methods. The dynamic power of gray level co-occurrence matrix(GLCM) and fractal dimension(FD) in the remote sensing cloud detection based on texture feature is evaluated using the proposed method. The average error of the dynamic power for GLCM and FD is 11.86% using the proposed method. The computing time for GLCM and FD using the proposed method is 0.295 times than the method based XPower.
收录类别:CPCI-S
资源类型:会议论文
TOP