多核系统中基于动态电压频率调节的实时节能调度研究  被引量:6

Research on Real-Time Energy-Efficient Scheduling in Multi-Core Systems Based on Dynamic Voltage Frequency Scaling

在线阅读下载全文

作  者:张冬松[1] 陈芳园[1] 金士尧[1] 

机构地区:[1]并行与分布处理国防科技重点实验室,湖南长沙410073

出  处:《计算机工程与科学》2010年第9期157-164,共8页Computer Engineering & Science

基  金:国家教育部博士点基金资助项目(20079998014);国家自然科学基金资助项目(60773024)

摘  要:在实时嵌入式领域,特别是无线移动和便携式计算领域,能耗是首要考虑的因素,这也是多核处理器尚未在嵌入式领域全面展开应用的首要因素。目前针对多核系统的实时应用,基于动态电压频率调节(DVFS)的实时节能调度技术研究得较少,还有许多问题亟待解决。本文介绍了多核系统中动态电压频率调节技术,分析讨论了当前多核系统中实时调度研究进展,主要针对同构多核、异构多核、并行任务模型和弱硬实时模型等方面,深入探讨了多核系统中基于DVFS的实时节能调度。本文结合多核系统、电压频率动态调节节能和实时调度,探索了多核系统中的实时节能调度,奠定了理论和技术基础,具有重大的理论意义和现实应用价值。In the real-time embedded areas,especially in wireless mobile and portable computing,energy consumption is now of primary consideration,which is the primary reason that the applications of multi-core processors are still in full swing for the embedded field. As to real-time applications for the current multi-core systems,a real-time scheduling technology based on Dynamic Voltage Frequency Scaling (DVFS) is studied less,and there are also many scientific and technical issues to be resolved. This paper introduces dynamic voltage frequency scaling in the multi-core systems,analyses and discusses the research evolvement of real-time scheduling in the current multi-core systems,then goes deep into DVFS-based real-time energy-efficient scheduling in the multi-core systems mainly for the aspects of homogeneous multi-core,heterogeneous multi-core,parallel task model and weakly hard real-time model. Combining multi-core systems,DVFS-based energy management and real-time scheduling to explore the real-time energy-efficient scheduling in the multi-core systems,laying a theoretical and technical basis,the paper is of great theoretical significance and practical application importance.

关 键 词:实时系统 多核系统 动态电压频率调节 并行任务 弱硬实时 

分 类 号:TP316[自动化与计算机技术—计算机软件与理论]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象