A Novel Approach for Branch Buffer Consuming Power Reduction

B. Zamani, E. Adeli, H. Gharedaghi, M. Soryani
{"title":"A Novel Approach for Branch Buffer Consuming Power Reduction","authors":"B. Zamani, E. Adeli, H. Gharedaghi, M. Soryani","doi":"10.1109/ICCEE.2008.48","DOIUrl":null,"url":null,"abstract":"By increasing the pipeline length in processors, the accuracy of the branch predictor unit plays an important role in processors efficiencies. In addition to the accuracy, consuming power is also an essential in portable systems. Therefore, in the processors these days the prediction unit is used to determine the branch destination, while most of these accesses are not necessary. In this paper, a method is proposed to reduce the consuming power in the jump prediction unit. In this proposed method, the non-necessary accesses to BTB are reduced by taking into account this fact that there exists distances between different consecutive branch instructions. This method decides the access to BTB by a constant value and a counter. After an instruction entrance, the BTB is accessed if the counter is zero, and if the instruction is a branch instruction and exists in the BTB the counter is reset. The simulation and experimental results illustrate the suitable performance of the proposed method in comparisons to the other methods. This superiority is for both the execution time and for the consuming power. Also it is more strengthened by increasing the distance.","PeriodicalId":365473,"journal":{"name":"2008 International Conference on Computer and Electrical Engineering","volume":"142 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2008-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2008 International Conference on Computer and Electrical Engineering","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCEE.2008.48","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2

Abstract

By increasing the pipeline length in processors, the accuracy of the branch predictor unit plays an important role in processors efficiencies. In addition to the accuracy, consuming power is also an essential in portable systems. Therefore, in the processors these days the prediction unit is used to determine the branch destination, while most of these accesses are not necessary. In this paper, a method is proposed to reduce the consuming power in the jump prediction unit. In this proposed method, the non-necessary accesses to BTB are reduced by taking into account this fact that there exists distances between different consecutive branch instructions. This method decides the access to BTB by a constant value and a counter. After an instruction entrance, the BTB is accessed if the counter is zero, and if the instruction is a branch instruction and exists in the BTB the counter is reset. The simulation and experimental results illustrate the suitable performance of the proposed method in comparisons to the other methods. This superiority is for both the execution time and for the consuming power. Also it is more strengthened by increasing the distance.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
一种降低支路缓冲器功耗的新方法
通过增加处理器的管道长度,分支预测单元的准确性对处理器的效率起着重要的作用。除了精度之外,在便携式系统中,功耗也是必不可少的。因此,在目前的处理器中,预测单元用于确定分支目的地,而大多数这些访问都是不必要的。本文提出了一种降低跳变预测单元功耗的方法。在该方法中,考虑到不同连续分支指令之间存在距离这一事实,减少了对BTB的不必要访问。这个方法通过一个常量和一个计数器来决定对BTB的访问。在指令进入后,如果计数器为零,则访问BTB,如果该指令是分支指令并且存在于BTB中,则重置计数器。仿真和实验结果表明,与其他方法相比,该方法具有良好的性能。这种优势既体现在执行时间上,也体现在消耗能力上。此外,它是更加强了增加距离。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
Understanding Code Patterns Analysis, Interpretation and Measurement Community Collaborative Filtering for E-Learning A Nonlinear Control Approach to Increase Power Oscillations Damping by SSSC Automatic Recognition of Pavement Surface Crack Based on BP Neural Network A Dynamic Model for Early Vision Processing
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1