期刊文献+

一种基于二步流控方法的片上动态虚通道路由器 被引量:1

An on-Chip Dynamic Virtual Channel Router Based on Two-Step Flow Control Method
下载PDF
导出
摘要 片上硅面积和功耗受到严重限制,报文缓冲区容量也受到严重限制,如何高效使用报文缓冲区是NoC设计的关键问题之一.动态划分虚通道缓冲区是高效使用报文缓冲区的有效方法之一,但会增加拥塞程度,甚至出现无限拥塞的情况.提出一种基于二步流控方法的片上动态虚通道(DAVC)路由器,该二步流控方法将报文分成报文头和报文体两部分分别运用流控算法.实验结果表明:与静态虚通道(SAVC)片上路由器相比,在缓存容量相等的情况下,DAVC路由器能提高23.2%的吞吐率,传输延迟降低27.2%;在DAVC缓存容量减半的情况下可获得相近的性能,节省28.3%的面积与23.8%的漏电流功耗. The on-chip area and power are limited seriously,so as to the capacity of packet buffer.So one of key issues of NoC design is how to use packet buffer efficiently.Dynamic virtual channel buffer is one of efficient ways,but may make congestion heavier,or even may make dead congestion appear.This paper proposes an on-chip dynamic virtual channel(DAVC) router based on two-step flow control method,which splits a packet into a head and a body to apply flow control algorithms separately,and it can retard congestion and stop dead congestion.There is a centralized share buffer on every input port in the router.The centralized share buffer is dynamic allocated to every packet of this input port according to the conditions of traffic and flow control,and can implement zero cycle delay between a read and a write and support variable packet length.The router also uses a two-level crossbar arbitrator and a novel virtual channel allocator based on tree,which allocates output virtual channel dynamically.The experiment results show that,compared with static virtual channel(SAVC),DAVC router provides 23.2% throughput improvement and 27.2% delay reduction with same buffer capacity,and provides 28.3% area and 23.8% power savings while having similar performance with half buffer capacity.
出处 《计算机研究与发展》 EI CSCD 北大核心 2011年第1期36-44,共9页 Journal of Computer Research and Development
基金 国家自然科学基金项目(60676010) 国家"八六三"高技术研究发展计划基金项目(2007AA01Z108) 教育部"长江学者和创新团队发展计划"基金项目
关键词 流控 动态虚通道 路由器 片上网络 缓冲区管理 flow control dynamic virtual channel router network-on-chip buffer management
  • 相关文献

参考文献3

二级参考文献32

  • 1Dally William J. Route packets, not wires: On-chip interconnection networks//Proceedings of the 38th Design Automation Conference. Las Vegas, NV, 2001:684-689
  • 2Karol M J et al. Input versus output queueing on a space-division packet switch. IEEE Transactions on Communications, 1987, 35(12): 1347-1356
  • 3Nikolay Kavaldjiev, Smit Gerard J M. A virtual channel network-on chip for GT and BE traffic//Proceedings of the VLSI Technologies and Architectures. Karlsruhe, Germany, 2006:211-216
  • 4Aline Mello, Leonel Tedesco. Virtual channel in network on chip: Implementation and evaluation on hermes NoC//Proceedings of the 18th Annual Symposium on Integrated Circuits and System Design. Florianopolis, Brazil, 2005: 178- 183
  • 5Mullins R, West A, Moore S. Low-latency virtual-channel routers for on-chip networks//Proceedings of the IEEE Symposium on Computer Architecture. Munchen, Germany, 2004:188-200
  • 6Wang H, Zhu X, Peh L, Malik S. Orion: A power-perform ance simulator for interconnection networks//Proceedings of the 35th Annual IEEE/ACM International Symposium on Microarchitecture. Istanbul, Turkey, 2002:294-305
  • 7Chen X, Peh L. Leakage power modeling and optimization in interconnection networks//Proceedings of the 2003 International Symposium on Low Power Electronics and Design. Seoul, Korea, 2002:90-95
  • 8Hu J, Ogras U Y. System-level buffer allocation for application-specific networks-on-chip router design. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2006, 25(12): 2919-2933
  • 9Huang Ting-Chun, Ogras Umit Y. Virtual channels planning for networks-on-chip//Proceedings of the 8th International Symposium on Quality Electronic Design. San Jose, CA, 2007: 879-884
  • 10Tamir Y, Frazier G L. High-performance multiqueue buffers for VLSI communication switches//Proceedings of the Annual International Symposium on Computer Architecture. Honolulu, USA, 1988: 343-354

共引文献18

同被引文献20

  • 1Benini L,Micheli G D. Networks on chips:A new soc paradigm[J].IEEEComputer,2002,(1):70-78.
  • 2Constantinescu C. Trends and challenges in VLSI circuit reliability[J].{H}IEEE MICRO,2003,(4):14-19.
  • 3Moscibroda T,Mutlu O. A case for bufferless routing in onchip networks[A].New York:ACM,2009.196-207.
  • 4Hayenga M,Jerger N E,Lipasti M. SCARAB:A single cycle adaptive routing and bufferless network[A].New York:ACM,2009.244-254.
  • 5Feng Chaochao,Lu Zhonghai,Jantsch A. A reconfigurable fault-tolerant deflection routing algorithm based on reinforcement learning for network-on-chip[A].New York:ACM,2010.11-16.
  • 6Dumitras T,Kerner S,Marculescu R. Towards on-chip fault-tolerant communication[A].New York:ACM,2003.225-232.
  • 7Pirretti M,Link G M,Brooks R R. Fault tolerant algorithms for network-on-chip interconnect[A].Los Alamitos,CA:IEEE Computer Society,2004.46-51.
  • 8Zhen Zhang,Greiner A,Taktak S. A reconfigurable routing algorithm for a fault tolerant 2D-mesh Network-on-Chip[A].New York:ACM,2008.441-446.
  • 9Fick D,DeOrio A,Chen G. A highly resilient routing algorithm for fault-tolerant NoCs[A].Leuven,Belgium:European Design and Automation Association,2009.21-26.
  • 10Lillis J,Cheng C. Timing optimization for multisource nets:Characterization and optimal repeater insertion[J].{H}IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems,1999,(3):322-331.

引证文献1

二级引证文献8

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