期刊文献+

网络并行超级计算系统THNPSC-1 被引量:3

Networked Parallel Supercomputing System THNPSC-1
下载PDF
导出
摘要 网络并行计算 (也称集群式计算 )是实现高性能计算的重要方式 ,该文介绍了一个清华大学研制的网络并行超级计算系统 THNPSC- 1.它是由 Pentium III SMP计算结点组成 ;网络互联采用两种高速网 :一种是自制的具有动态仲裁与路由寻经的交叉开关网络 THNet,另一种是 10 0 Mpbs的 Ethernet.THNet中的交叉开关 THSwitch是用 15万门的 AL TERA FPGA芯片构成 ,THNet还包括具有 DMA引擎的网络适配器 THNIA .THNet每一端口可以提供数据传输率为 1.0 5 6 Gbps,其聚合频宽可达 8.4Gbps;采用固定用户缓冲和扩展的主动消息传递等方法 ,THNet执行用户层的消息传递 ,旁路操作系统的系统调用 ,做到零拷贝的消息传递 .乒乓测试结果表明 :单向消息传递延迟可减少到 8μs.THNet软件包括 THNIA驱动程序和支持用户层通信的函数库 .此文对相关工作进行了简要对比 。 Networked parallel computing system NPCS (sometimes termed as cluster computing system) has become increasingly important in high performance computing. This paper describes a networked parallel supercomputer THNPSC 1 built at Tsinghua University. It consists of 16 dual pentium III SMP nodes. It employs two kinds of high speed interconnection networks. One is custom designed crossbar switch network THNet with dynamic arbitration and routing capability, built by Tsinghua University. The other is a 100Mbps Ethernet. The crossbar switch THSwitch of the THNet is built on the basis of Altera FPGA chips with 150,000 gates, and THNet also contains the network adapters THNIA. THNet is able to provide 1.056 Gbps data transmission rate for each port, and the aggregate bandwidth can reach 8.4 Gbps. By using the technique of fixed user message buffer management and extend active message [9] , THNet carried out the user level message passing, bypassing the operation system with zero data copy. The measurement indicates that communication latency of message passing can be reduced to 8 Microsecond. The software of THNet includes the driver of THNIA, and the library for supporting the user level communication. This paper also gives a comparison with the related work, and depicts the applications of THNPSC 1.
出处 《计算机学报》 EI CSCD 北大核心 2001年第6期627-632,共6页 Chinese Journal of Computers
基金 国家攀登计划 B高性能计算若干关键技术研究项目资助
关键词 集群式计算 网络并行超级计算系统 互联网 HNMPSC-1 networked parallel computing, cluster computing, fast crossbar switch network, user level communication, performance evaluation tool
  • 相关文献

参考文献3

二级参考文献6

  • 1郑耿斌.PACT HPF标准数据映射语句的定义.技术报告[M].,1996..
  • 2都志辉.HPF并行语句及结构的标量化算法.技术报告[M].,1996..
  • 3都志辉 汪剑平.PACT HPF子集的定义.技术报告[M].,1997..
  • 4都志辉 汪剑平 等.Laplace例子的测试结果.技术报告[M].,1997..
  • 5都志辉.HPF程序与手写结点程序的加速比.技术报告[M].,1997..
  • 6Chang Sheueling,IEEE Parrallel & Distributed Technology,1995年,20页

共引文献16

同被引文献9

  • 1RajkumarBuyya 郑纬民 等.高性能集群系统(一)、(二)卷[M].北京:电子工业出版社,2001.5-6.
  • 2Duato J,Yalamanchili S,Ni L.并行计算机互连网络技术:一种工程方法[M].谢伦国,张民选,窦强,等,译.北京:电子工业出版社,2004.
  • 3Agarwal A,Bianchini R,et al.The MIT alewife machine[J].Proceedings of the IEEE,1999,87(3):430-444.
  • 4Intel corporation.Paragon XP/s product overview[R].Beaverton,Oregon,USA:Supercomputer Systems Division,1991.
  • 5Chiara Francalanci,Paolo Giacomazze.High-performance self-routing algorithm for multiprocessor systems with shuffle interconnections[J].IEEE Transactions on Parallel and Distributed Systems,2006,17(1):38-50.
  • 6Qiao Baojun,Shi Feng,Ji Weixing.A new routing algorithm in triple-based hierarchical interconnection network[C].Beijing:First International Conference on Innovative Computing,Information and Control,2006:725-728.
  • 7Dianne R Kumar,Walid A Najjar,Pradip K Srimani.A new adaptive hardware tree-based multicast routing in K-ary n-cubes[J].IEEE Transactions on Computers,2001,50(7):647-658.
  • 8杜毅,李三立.k-ary n-cube网络中高速开关TH-Switch的设计与路由算法[J].计算机学报,1999,22(1):16-23. 被引量:5
  • 9戴梅萼,张灏,周桂林.高性能计算机系统接口通信机制的研究[J].小型微型计算机系统,2001,22(3):278-279. 被引量:1

引证文献3

二级引证文献3

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部