A Storage Structure Design Based on the Cache Mechanism in a Multi-core SoC

ZHANG Duo-li; ZHANG Yu; SONG Yu-kun; WANG Jian

Chinese

您当前的位置：

首页 >

文章列表页 >

A Storage Structure Design Based on the Cache Mechanism in a Multi-core SoC

更新时间：2026-09-04

- A Storage Structure Design Based on the Cache Mechanism in a Multi-core SoC
- Vol. 34, Issue 10, Pages: 26-31(2017)
- 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：2017
- 稿件说明：
移动端阅览
张多利, 张宇, 宋宇鲲, 汪健. 一种多核SoC中基于Cache机制的存储结构设计[J]. 微电子学与计算机, 2017, 34(10): 26-31.

ZHANG Duo-li, ZHANG Yu, SONG Yu-kun, et al. A Storage Structure Design Based on the Cache Mechanism in a Multi-core SoC[J]. 2017, 34(10): 26-31.
张多利, 张宇, 宋宇鲲, 汪健. 一种多核SoC中基于Cache机制的存储结构设计[J]. 微电子学与计算机, 2017, 34(10): 26-31. DOI：

ZHANG Duo-li, ZHANG Yu, SONG Yu-kun, et al. A Storage Structure Design Based on the Cache Mechanism in a Multi-core SoC[J]. 2017, 34(10): 26-31. DOI：

摘要

本文针对一种面向高密度计算的异构多核SoC系统

提出了一种层次化的共享二级存储结构(L2-Cache)

以缓解系统数据处理速度与外部存储间的速度差异.所设计的层次化存储结构提供对象数据缓存功能

利用计数替换策略

减少二级存储污染

提高有效数据命中率; 在计算时间间隙实现数据准确预读取和L2-主存同步操作

增加有效存储带宽.最终测试结果表明

采用层次化存储结构的设计兼顾了不同访存比应用的数据访存特性

平均访存性能提高31.1%

不同规模的矩阵运算最高获得1.573的加速比

整体任务计算时间平均减少了27.8%.

Abstract

A hierarchical shared secondary storage (L2-Cache) structure for a heterogeneous multi-core SoC face to high density calculation is proposed in this paper to ease the speed difference between data processing and data storage. The hierarchical storage structure provides object data caching function and reduce the pollution of secondary storage using count replacement strategy to improve valid data hit rate. The structure can also increase the effective memory bandwidth by L2-main memory synchronous operation and accurate prefetching in data computing interval time. Final test results show that

L2-Cache design can adapt for various datas feature of different applications.The new design increase the average fetch performance of about 31.1%

and gain 1.573 speed ratio highest of different sizes of matrix

and reduce the average computing time of about 27.8%.

关键词

Keywords

references

Joseph D, Grunwald Dirk. Prefetching using markov predictors[J]. IEEE Transactions on Computers, 1999, 48(2): 121-133. DOI: 121

李柱. 分布式文件系统小文件性能优化技术研究与实现[D]. 长沙: 国防科学技术大学, 2008.

肖瑞瑾. 多核处理器层次化存储体系研究[D]. 上海: 复旦大学, 2012.

董立平, 周丽. 多核多线程是未来处理器的发展方向[J]. 高性能计算发展与应用, 2008, 6(3): 29-32. DOI: 29

Wang Feng. Storage management in large distributed object-based storage system[J]. Journal of Parallel % Distributed Computing, 2009, 69(4): 349-359 DOI: 349

晏沛湘. 片上多核处理器二级Cache结构及资源管理技术研究[D]. 长沙: 国防科学技术大学, 2012.

Vander S P, Lilja D J. Data prefetch mechanisms[J]. ACM Computing Surveys, 2000, 32(2): 174-199. DOI: 174

孙玲. 多通道并行访存接口设计与实现[D]. 合肥: 合肥工业大学, 2015.

朱旭东, 柯剑, 那文武, 等. 一种数据块关系指导的缓存替换方法[C]//第15届全国信息存储技术学术会议论文集. 北京, 2008: 8.

Megiddo N, Modha D. ARCl A self-tuning, low overhead replacement cachel[C]//Proc of the 2nd USENIX Syrup on File and Storage Technologies. USA, San Francisico, 2003: 115-130

王庆. 面向嵌入式多核系统的并行程序优化技术研究[D]. 哈尔滨: 哈尔滨工业大学, 2013.

Lee D, Choi J, Kim J H, et al. LRFU: A spectrum of policies that subsumes the least recently used and least frequently used policies[J]. IEEE Trans on Computers, 2001, 50(12): 1352-1360 DOI: 1352

Netherland John L Hennessy, David A Patterson. Computer achitecture: a quantitative approach[M]. 4th ed. Netherland, Elsevier Inc, 2007.

谢学军. 数据值预测与存储器预取技术研究[D]. 哈尔滨: 哈尔滨工业大学, 2006.

孙国忠, 袁清波, 陈明宇, 等. 用于二级缓存的一种改进的自适应缓存管理算法[J]. 计算机研究与发展, 2007, 49(8): 1331-1338. DOI: 1331

方娟, 张红波. 一种改进的多核处理器硬件预取技术[J]. 计算机科学, 2012, 39(10): 48-50. DOI: 48

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Simulation on the DC characteristics of double heterojunction p-GaN HEMT with AlN/GaN hybrid buffer layer

A feature-oriented context understanding and slice analysis network for tuberculosis detection

A high-efficiency charge pump for ultra-low voltage boost systems

Monte carlo analysis of V-layout trench SiC neutron detector

Design of ultra-low power real-time clock circuits

Related Author

No data

Related Institution

No data

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