A Prediction Method for Upper Bound of Reorder Buffer of NoCs Based on Network Calculus

DU Gao-ming; SONG Ping; SONG Yu-kun; ZHANG Duo-li; LI Miao

Chinese

您当前的位置：

首页 >

文章列表页 >

A Prediction Method for Upper Bound of Reorder Buffer of NoCs Based on Network Calculus

更新时间：2026-09-04

- A Prediction Method for Upper Bound of Reorder Buffer of NoCs Based on Network Calculus
- Vol. 33, Issue 4, Pages: 6-9(2016)
- 作者机构：
  
  合肥工业大学微电子设计研究所,安徽,合肥,230009
- 作者简介：
- 基金信息：
- DOI：
  CLC：
- Published：2016
- 稿件说明：
移动端阅览
杜高明, 宋平, 宋宇鲲, 张多利, 李苗. 基于网络演算的片上网络重组缓存上界的预测方法[J]. 微电子学与计算机, 2016, 33(4): 6-9, 15.

DU Gao-ming, SONG Ping, SONG Yu-kun, et al. A Prediction Method for Upper Bound of Reorder Buffer of NoCs Based on Network Calculus[J]. 2016, 33(4): 6-9.
杜高明, 宋平, 宋宇鲲, 张多利, 李苗. 基于网络演算的片上网络重组缓存上界的预测方法[J]. 微电子学与计算机, 2016, 33(4): 6-9, 15. DOI：

DU Gao-ming, SONG Ping, SONG Yu-kun, et al. A Prediction Method for Upper Bound of Reorder Buffer of NoCs Based on Network Calculus[J]. 2016, 33(4): 6-9. DOI：

摘要

多路径路由片上网络(Network on Chip

NoC)中

业务流之间的冲突导致数据包乱序到达目的节点

在目的节点处的重组缓存通过暂存乱序到达的数据包以保证数据包按照发包的先后顺序输出.传统方法通过全遍历所有路径得到重组缓存的最大值

即重组缓存上界

对于较大规模的网络

全遍历耗时长且效率低.对此提出一种得到重组缓存上界的预测方法: 通过冲突矩阵

推导出每条目标子业务流的冲突系数

进而选出冲突系数最大的无交叉路径组

通过路径组来预测重组缓存上界.相比于全遍历方法

该预测方法可以缩短分析时间

简化算法复杂度

提高效率

从而有效减少重组缓存区面积

降低片上网络设计成本.

Abstract

In multi-path routing NoCs

the problem of packets disorder will occur due to the conflicts among flows. Reorder buffer at target node can keep the order of sending the packets by storing the packets temporarily. Traditional way is traversing all the paths to get the maximal value of reorder buffer

which is the upper bound of reorder buffer. As to a network of lager scale

traversal way is a time-consuming and low efficiency process. An prediction method to calculate the upper bound of reorder buffer was proposed: derive conflict coefficient of each target sub-flow from conflict matrix

and select the non-cross path pairs which have maximal conflict coefficient

predict the upper bound of reorder buffer through these path pairs. Compared to traditional way

this prediction method can shorten analysis time

simplify algorithm complexity

improve efficiency

and thus effectively reduce reorder buffer area and the design cost of NoC.

关键词

Keywords

references

Du G, Zhang C, Lu Z, et al. Worst-case performance analysis of 2-D mesh NoCs using multi-path minimal routing[C]//Proceedings of the eighth IEEE/ACM/IFIP international conference on Hardware/software codesign and system synthesis. New York: ACM, 2012: 123-132.

Kwon W C, Yoo S, Hong S M, et al. A practical approach of memory access parallelization to exploit multiple off-chip DDR memories[C]//Design Automation Conference, 2008, 45th ACM/IEEE. Anaheim: IEEE, 2008: 447-452.

Du G, Li M, Lu Z, et al. An analytical model for worst-case reorder buffer size of multi-path minimal routing NoCs[C]//Networks-on-Chip (NoCS), 2014 Eighth IEEE/ACM International Symposium on. Ferrara: IEEE, 2014: 49-56.

Ebrahimi M, Daneshtalab M, Liljeberg P, et al. A high-performance network interface architecture for NoCs using reorder buffer sharing[C]//Parallel, Distributed and Network-Based Processing (PDP), 2010 18th Euromicro International Conference on. Pisa: IEEE, 2010: 546-550.

Kwon W C, Yoo S, Um J, et al. In-network reorder buffer to improve overall NoC performance while resolvingthe in-order requirement problem[C]//Design, Automation % Test in Europe Conference % Exhibition, 2009. Nice: IEEE, 2009: 1058-1063.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Design and implementation of network interface for network on chip and DDR3 controller

Design of NoC router based on advance congestion-aware

Low Power Associated Coding Design Based on Multi-Path Routing NoC

Design of Network Interface Based on Mirco-code Controller

VLSI Design of Router for Network on Chip

Related Author

CHEN Junru

LI Qingxin

TIAN Qing

GAO Ying

JU Hu

WEI Jinghe

HAN Yujie

ZHANG Duo-li

Related Institution

The No.58 Research Institute of China Electronics Technology Group Corporation

National Key Laboratory of Integrated Circuits and Microsystems

Institute of Microelectronics and VLSI Design, Hefei University of Technology

School of Electionics Science&Applied Physics, Hefei University of Technology

Institute of VLSI Design, Nanjing University

AI问答

Postal code：100079
Tel：（010）53879206 Email：tmw@bjxintong.com.cn
Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09082226号-64 京公网安备11010602201714号
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰