A Power-Aware Dead Lock Avoid Three-Dimensional Full-Adaptive Routing Algorithm for 3D NoC

YU Xiao; LI Li; ZHANG Yu-ang; PAN Hong-bing; WANG Jia-wen; HAN Ping

doi:10.3969/j.issn.0372-2112.2013.02.019

您当前的位置：

首页 >

文章列表页 >

A Power-Aware Dead Lock Avoid Three-Dimensional Full-Adaptive Routing Algorithm for 3D NoC

更新时间：2025-07-16

- A Power-Aware Dead Lock Avoid Three-Dimensional Full-Adaptive Routing Algorithm for 3D NoC
- Acta Electronica Sinica Vol. 41, Issue 2, Pages: 329-334(2013)
- 作者机构：
  
  南京大学电子科学与工程学院,江苏,南京,210093
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China (No.61176024, No.60876017)
- DOI：10.3969/j.issn.0372-2112.2013.02.019
  CLC： TP391.7
- Published：2013
- 稿件说明：
移动端阅览
YU Xiao, LI Li, ZHANG Yu-ang, et al. A Power-Aware Dead Lock Avoid Three-Dimensional Full-Adaptive Routing Algorithm for 3D NoC[J]. Acta Electronica Sinica, 2013, 41(2): 329-334.
DOI：

YU Xiao, LI Li, ZHANG Yu-ang, et al. A Power-Aware Dead Lock Avoid Three-Dimensional Full-Adaptive Routing Algorithm for 3D NoC[J]. Acta Electronica Sinica, 2013, 41(2): 329-334. DOI： 10.3969/j.issn.0372-2112.2013.02.019.

摘要

随着近年来三维片上网络(3D NoC)技术的提出及不断发展

功耗问题已成为3D NoC设计中面临的严峻挑战之一.本文为3D NoC提出一种面向功耗免死锁三维全动态路由算法TFRA (Three-dimensional Full-adaptive Routing Algorithm).其以传统二维NoC奇偶拐弯模型为基础

将三维路由空间划分为8个象限

针对每个象限制定相应的路由策略

从而实现免死锁.采用SystemC系统级建模语言搭建的3D NoC仿真平台进行验证

结果显示TFRA算法在功耗性能指标方面较现有的三维路由算法有大幅提升.

Abstract

With the proposal and development of the technology of 3D NoC in these years

the power issue has become one of the serious challenges in the design of 3D NoC.This paper proposed a power-aware deadlock avoid three-dimensional full-adaptive routing algorithm called TFRA for 3D NoC.The new routing strategy is based on the even-odd turning model of the conventional 2D NoC.The routing space is divided into eight different quadrants in which corresponding routing strategy is used.We use a systematic modeling language SystemC to build the platform of 3D NoC to test and verify the performance.The result shows that great improvement can be achieved considering every kind of index of power performance using our TFRA when compared to the traditional 3D routing algorithm.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Graph Learning Based Transmission Resources Scheduling in Dense Space Networks

New Theories and Methods of Image Segmentation

Multi-hop WLAN Video Distortion Optimized Routing and Source Rate Control

Design of Multicast Network of Dawning 5000 High Performance Computer

A Contention-Based Forwarding Routing Protocol for Vehicular Ad Hoc Networks in City Scenarios LI Yuan-zhen^1,2,LIAO Jian-xin^1,2,LI Tong-hong³,ZHU Xiao-min1,2

Related Author

LIU Run-zi

WU Wei-hua

ZHANG Wen-zhu

ZHOU Di

ZHANG Yan

FONT face

Verdana

XU Xin-zheng

Related Institution

The State Key Laboratory of ISN， Xidian University

School of Information and Control Engineering， Xi’an University of Architecture and Technology

中国科学院计算技术研究所智能信息处理重点实验室

1．中国矿业大学计算机科学与技术学院

（北京交通大学电子信息工程学院，）

⁰