全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...
电子学报  2012 

一种降低末级高速缓存污染的软件控制插入策略

, PP. 2433-2438

Keywords: 末级高速缓存,剖视,插入策略

Full-Text   Cite this paper   Add to My Lib

Abstract:

现有高速缓存替换算法大多无法有效识别数据的局部性特征,导致高速缓存内即将被访问到的数据可能被未来不会被访问到的数据所替换,造成高速缓存污染问题.末级高速缓存污染引发的性能损失随着处理器和存储器之间性能差距的扩大而不断增大,成为制约系统性能提升的重要瓶颈之一.本文针对末级高速缓存污染问题,在剖视分析访存密集型程序的页一级访存行为基础上,提出一种软件控制末级高速缓存插入策略.本方法通过控制和指导页一级数据插入位置,限制局部性差的数据页在末级高速缓存中的访问空间,达到降低末级高速缓存污染的目的.实验结果表明,相对于LRU和DIP策略,本文方法能够有效降低末级高速缓存失效率,提高程序性能.

References

[1]  S Livio,et al.Reducing the harmful effects of last-level cache polluters with an OS-level,software-only pollute buffer .Int’l Symposium on Microarchitecture .Washington:IEEE Computer Society,2008.258-269.
[2]  K Mazen,et al. Counter-based cache replacement algorithms .Int’l Conference on Computer Design .Washington:IEEE Computer Society,2005.61-68.
[3]  R E Kessler,et al.Page placement algorithms for large real-indexed caches[J].ACM Transactions on Computer Systems,1992,10(4):338-359.
[4]  L Qingda,et al.Soft-OLP:Improving hardware cache performance through software-controlled object-Level partitioning .Int’l Conference on Parallel Architectures and Compilation Techniques .Washington:IEEE Computer Society,2009.246-257.
[5]  L H John,et al.Spec CPU2000:Measuring CPU performance in the new millennium[J].Washington:IEEE Computer Society Press,2000,33(7):28-35.
[6]  K Q Moinuddin,et al.Adaptive insertion policies for high performance caching .Int’l Symposium on Computer Architecture .New York:ACM SIGARCH Computer Architecture News,2007.381-391.
[7]  D Haakon,et al.An LRU-based replacement algorithm augmented with frequency of access in shared chip-multiprocessor caches .Int’l Workshop on Memory Performance:Dealing with Applications,Systems and Architectures .New York:ACM,2007.45-52.
[8]  L Jiang,et al.Enabling software management for multicore caches with a lightweight hardware support .Int’l Conference on High Performance Computing Networking,Storage and Analysis .New York:ACM,2009.1-12.
[9]  G Xiaoming,et al.On the theory and potential of LRU-MRU collaborative cache management .Int’l Symposium on Memory Management .New York:ACM,2011.43-54.
[10]  S Andreas,et al.Reducing cache pollution through detection and elimination of non-temporal memory accesses .Int’l Conference on High Performance Computing,Networking,Storage and Analysis .Washington:IEEE Computer Society,2010.1-11.
[11]  邓让钰,陈海燕,等.一种异构多核处理器的并行流存储结构[J].电子学报,2009,37(2):312-317. Deng Rangyu,Chen Haiyan,et al.A parallel stream memory architecture for heterogeneous multi-core processor[J].Acta Electronica sinica,2009,37(2):312-317.(in Chinese)
[12]  J Aamer,et al.High performance cache replacement using re-reference interval prediction (RRIP) .Int’l Symposium on High-Performance Computer Architecture .New York:ACM,2010.60-71.
[13]  Z Xiao,et al.Towards practical page coloring-based multicore cache management .Int’l Conference on Computer Systems of European .New York:ACM,2009.89-102.
[14]  B Doug,et al.The SimpleScalar tool set,version 2.0[J].ACM SIGARCH Computer Architecture News,1997,25(3):13-25.
[15]  C Dhruba,et al.Predicting inter-thread cache contention on a chip multi-processor architecture .Int’l Symposium on High-Performance Computer Architecture .Washington:IEEE Computer Society,2005.340-351.
[16]  P Erez,et al.Using SimPoint for accurate and efficient simulation .Int’l Conference on Measurement and Modeling of Computer Systems .New York:ACM,2003.318-319.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133