3d International ICST Conference on Bio-Inspired Models of Network, Information, and Computing Systems

Research Article

A Swarm-Inspired Resource Distribution for SMT Processors

Download532 downloads
  • @INPROCEEDINGS{10.4108/ICST.BIONETICS2008.4694,
        author={Hongzhou Chen and Lingdi Ping and Xuezeng Pan and Kuijun Lu and Xiaoning Jiang},
        title={A Swarm-Inspired Resource Distribution for SMT Processors},
        proceedings={3d International ICST Conference on Bio-Inspired Models of Network, Information, and Computing Systems},
        publisher={ICST},
        proceedings_a={BIONETICS},
        year={2010},
        month={5},
        keywords={Simultaneous multithreading resource distribution bio-inspired model optimization method swarm intelligence},
        doi={10.4108/ICST.BIONETICS2008.4694}
    }
    
  • Hongzhou Chen
    Lingdi Ping
    Xuezeng Pan
    Kuijun Lu
    Xiaoning Jiang
    Year: 2010
    A Swarm-Inspired Resource Distribution for SMT Processors
    BIONETICS
    ICST
    DOI: 10.4108/ICST.BIONETICS2008.4694
Hongzhou Chen1,*, Lingdi Ping1, Xuezeng Pan1, Kuijun Lu1, Xiaoning Jiang2,*
  • 1: College of Computer Science, Zhejiang University, Hangzhou, China.
  • 2: Sunyard System Engineering Co. Ltd., Hangzhou, China
*Contact email: honjoychan@gmail.com, jxn@sunyard.com

Abstract

The performance in Simultaneous Multi-Threading (SMT) processors is mainly determined by the distribution of the common resources among the threads. However, the threads exhibit dynamically complicated behavior while they compete for resources at runtime. It is a challenge to meet the changing resource requirements of the threads. This work proposes a Swarm-inspired Resource Distribution (SRD) policy to address the dynamic optimization problem of resource distribution for SMT processors, which uses the runtime performance to guide the generating of trial distributions. A computational model is established by adaptation of swarm intelligence to direct the social exploitation and self exploration activities of the trial distributions in the dynamic optimization environment. Results from simulation show that, benefiting from the good cooperation between SRD’s social exploitation on historical experience and self exploration of new solutions, SRD obtains satisfying improvements of both throughput and fairness performance, especially in complicated SMT environment.