基于SimpleScalar的龙芯CPU模拟器Sim-Godson  被引量:25

Sim-Godson:A Godson Processor Simulator Based on SimpleScalar

在线阅读下载全文

作  者:张福新[1] 章隆兵[1] 胡伟武[1] 

机构地区:[1]中国科学院计算技术研究所系统结构重点实验室,北京100080

出  处:《计算机学报》2007年第1期68-73,共6页Chinese Journal of Computers

基  金:国家杰出青年基金(60325205);国家自然科学基金(60673146);中国科学院计算技术研究所基础研究基金(20056020);中国科学院计算技术研究所知识创新课题(20056240);国家"八六三"高技术研究发展计划项目基金(2002AA110010;2005AA110010;2005AA119020);国家"九七三"重点基础研究发展规划项目基金(2005CB321600)资助.

摘  要:现代高性能通用处理器的设计越来越复杂,模拟器在处理器设计中所起的作用越来越大.龙芯2号是中国科学院计算技术研究所研制的高性能通用处理器.最早开发的龙芯2号的模拟器ICT-Godson是信号级模拟器,它模拟了处理器的所有细节,十分准确,但速度和灵活性有较大限制.文章基于SimpleScalar工具集,设计并实现了龙芯2号的模拟器Sim-Godson.Sim-Godson具有高速度和高灵活性的优点,且准确性也很高.在3.0GHz的Pentium4微机上,Sim-Godson速度约为500K指令/s.大部份测试程序在Sim-Godson上的IPC(Instruction PerCycle)与ICT-Godson相差不到5%,达到了很高的准确性.Sim-Godson在龙芯2号的性能分析工作中发挥了重要作用.As the increase in complexity of modern processor design, simulator is becoming more and more important in processor design. Godson-2 is a high performance general purpose processor developed by the Institute of Computing Technology, Chinese Academy of Sciences. ICTGodson, the initial simulator of Godson-2, is a signal level simulator, which simulates all hardware details of Godson-2 and is very accurate. But the speed of ICT-Godson is not high enough and the flexibility is not good. Based on SimpleScalar tool sets, Sire-Godson, one processor sireulator of Godson-2, is son has high speed and designed and implemented. Comparing with the ICT-Godson, Sire-Godhigh flexibility and the accuracy is also very good. On PC of Pentium4 3.0GHz, the speed of Sim-Godson is about 500K instructions per second. IPC(Instruction Per Cycle) difference between Sim-Godson and ICT-Godson is less than 5 percent for the most applications. Sim-Godson plays an important role in the performance analysis of God-son-2.

关 键 词:模拟器 龙芯2号处理器 SIMPLESCALAR 通用处理器 高性能处理器 

分 类 号:TP302[自动化与计算机技术—计算机系统结构]

 

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