基于Fat-tree的高性能互联网络性能优化与分析

张毅; 何卫锋

Chinese

您当前的位置：

首页 >

文章列表页 >

基于Fat-tree的高性能互联网络性能优化与分析

更新时间：2026-09-04

- 基于Fat-tree的高性能互联网络性能优化与分析
- Optimization and Evaluation of Fat-tree Based High Performance Interconnect Network
- 微电子学与计算机 2018年35卷第8期页码：36-41
- 作者机构：
  
  上海交通大学微电子学院,上海,200240
- 作者简介：
- 基金信息：
- DOI：
  中图分类号：
- 纸质出版：2018
- 稿件说明：
移动端阅览
张毅, 何卫锋. 基于Fat-tree的高性能互联网络性能优化与分析[J]. 微电子学与计算机, 2018, 35(8): 36-41.

ZHANG Yi, HE Wei-feng. Optimization and Evaluation of Fat-tree Based High Performance Interconnect Network[J]. 2018, 35(8): 36-41.
张毅, 何卫锋. 基于Fat-tree的高性能互联网络性能优化与分析[J]. 微电子学与计算机, 2018, 35(8): 36-41. DOI：

ZHANG Yi, HE Wei-feng. Optimization and Evaluation of Fat-tree Based High Performance Interconnect Network[J]. 2018, 35(8): 36-41. DOI：

摘要

高性能互联网络对于高性能计算机的性能具有重要的影响

而高性能网络中各个模块的设计都会影响到网络的最终性能

因而前期的网络仿真显得尤为重要.HPCSim是基于OMNeT++平台自主开发的高性能网络仿真器

包括了网络拓扑和仿真参数自动生成工具PySim以及仿真结果分析和可视化工具PyVisual.仿真器中对全缓存高阶路由器结构进行了改进

提出了一种5级流水线的高阶路由器

能够减少路由信息缓存开销同时将网络吞吐率从基准路由器下的0.75提升至0.95.此外

还对硅光互联链路进行了逻辑抽象

并在Fat-tree拓扑上对链路性能和网络吞吐率的关系进行了分析.

Abstract

High performance interconnect network has a significant influence on the performance of high performance computer. As the difference in network model design will result in different network performance

the early network simulation is needed. HPCSim is an independently developed high performance interconnect network simulator based on OMNeT++. It includes topology and parameter file generator named PySim

and simulation data analyzer and visualizer named PyVisual. An improved fully-buffered router structure is proposed

which achieves 5 stage pipeline

and reduces routing computation states buffer as well as promotes throughput from 0.75(baseline router) to 0.95. Moreover

a link model is inferred from photonic network

and the link performance simulation is run based on Fat-tree network to analyze the relation between link bandwidth and network throughput.

关键词

Keywords

references

Kim J, Dally W J, Towles B, et al. Microarchitecture of a high radix router[C]//Computer Architecture, 2005. ISCA'05. Proceedings. 32nd International Symposium on. Dresden, Germany, IEEE, 2005: 420-431.

Fu H, Liao J, Yang J, et al. The sunway taihu light supercomputer: system and applications[J]. Science China Information Sciences, 2016, 59(7): 072001. DOI: 072001

Pongor G. Omnet: Objective modular network testbed[C]//Proceedings of the International Workshop on Modeling, Analysis, and Simulation On Computer and Telecommunication Systems. Erice, Italy, Society for Computer Simulation International, 1993: 323-326.

Fulgham M L, Snyder L. Performance of chaos and oblivious routers under non-uniform tra c[R]. University of Washington, Seattle WA, Technical Report UW-CSE-93-06-01, 1993.

Hwang K, Wang C, Wang C L. Evaluating MPI collective communication on the SP2, T3D, and Paragon multicomputers[C]//High-Performance Computer Architecture, Third International Symposium on. San Antonio, Texas, IEEE, 1997: 106-115.

Liao X K, Pang Z B, Wang K F, et al. High performance interconnect network for Tianhe system[J]. Journal of Computer Science and Technology, 2015, 30(2): 259. DOI: 259

Kung N T, Morris R. Credit-based flow control for ATM networks[J]. IEEE network, 1995, 9(2): 40-48. DOI: 40

Leiserson C E. Fat-trees: universal networks for hardware-efficient supercomputing[J]. IEEE transactions on Computers, 1985, 100(10): 892-901. DOI: 892

Scott S, Thorson G. Optimized routing in the Cray T3D[J]. Parallel Computer Routing and Communication, 1994(853): 281-294. DOI: 281

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于区块链的AI生成内容溯源框架设计研究

大模型发展综述：从BERT到Deepseek的技术演进

一种用于极低压boost系统的高效率电荷泵

V型沟槽SiC中子探测器的蒙特卡洛分析

注意力机制与改进密集卷积引导的信息隐藏模型