• 综合
  • 标题
  • 关键词
  • 摘要
  • 学者
  • 期刊-刊名
  • 期刊-ISSN
  • 会议名称
搜索

作者:

Fang, Juan (Fang, Juan.) (学者:方娟) | Du, Wenjuan (Du, Wenjuan.) | Guo, Mei (Guo, Mei.) | Lei, Ding (Lei, Ding.) | Chen, Tan (Chen, Tan.)

收录:

EI Scopus

摘要:

This paper proposed a low-power-oriented cache design (LPD) for L2 cache of multi-core processors. We developed two methods to lower cache power consumption: LPHP (low power-oriented hybrid cache partition algorithm) and WPP-L2 (way-prediction based on a cache partition algorithm). The LPD considers these two different ways to reduce power consumption meanwhile to produce the best possible cache performance. In LPHP, we closed the columns that are not in use. In WPP-L2, we predicted an appropriate way before cache access that can reduce the access time to save power. Experiments show that LPHP can reduce power consumption by 23.7% and WPP-L2 can reduce 17.2% on average. In comparison with the traditional LRU strategy, these algorithms have little effect on the cache performance and sometimes improve the throughput. © 2015, by Binary Information Press

关键词:

Algorithms Cache memory Electric power supplies to apparatus Electric power utilization Integrated circuit design Microprocessor chips

作者机构:

  • [ 1 ] [Fang, Juan]College of Computer Science, Beijing University of Technology, Beijing, China
  • [ 2 ] [Du, Wenjuan]College of Computer Science, Beijing University of Technology, Beijing, China
  • [ 3 ] [Guo, Mei]College of Computer Science, Beijing University of Technology, Beijing, China
  • [ 4 ] [Lei, Ding]College of Computer Science, Beijing University of Technology, Beijing, China
  • [ 5 ] [Chen, Tan]College of Computer Science, Beijing University of Technology, Beijing, China

通讯作者信息:

  • 方娟

    [fang, juan]college of computer science, beijing university of technology, beijing, china

电子邮件地址:

查看成果更多字段

相关关键词:

相关文章:

来源 :

Journal of Computational Information Systems

ISSN: 1553-9105

年份: 2015

期: 17

卷: 11

页码: 6411-6418

被引次数:

WoS核心集被引频次: 0

SCOPUS被引频次:

ESI高被引论文在榜: 0 展开所有

万方被引频次:

中文被引频次:

近30日浏览量: 2

在线人数/总访问数:252/2892052
地址:北京工业大学图书馆(北京市朝阳区平乐园100号 邮编:100124) 联系我们:010-67392185
版权所有:北京工业大学图书馆 站点建设与维护:北京爱琴海乐之技术有限公司