Vol.14, NO.4, 2003
Turn off MathJax
Article Contents
Article Navigation >  Journal of Applied Meteorological Science  > 2003  >  14(4): 410-418
Jin Zhiyan, Wang Dingxing. An algorithmfor load balancing in a heterogeneous system. J Appl Meteor Sci, 2003, 14(4): 410-418.
Citation: Jin Zhiyan, Wang Dingxing. An algorithmfor load balancing in a heterogeneous system. J Appl Meteor Sci, 2003, 14(4): 410-418.
shu

AN ALGORITHMFOR LOAD BALANCING IN A HETEROGENEOUS SYSTEM

  • 1.

    Chinese Academy of Meteorological Sciences, Beijing 100081

  • 2.

    Department of Computer Science, Tsinghua University, Beijing 100084

  • Received Date: 2002-02-26
  • Rev Recd Date: 2002-06-25
  • Publish Date: 2003-08-31
    Abstract
  • Load balancing is a crucial problem in a heterogeneous system such as PC or workstation clusters, which has been widely used in the research and development of numerical weather prediction. A load balancing algorithm, based on feedback, is presented to eliminate the influence of uncertainty of processor speed, which is difficult to measure precisely. The influences of variation of processor speed, load distribution, errors of tested processor speed, etc., have been calculated. The results show that this method is quite robust. The diffusion equation and the simulated physical processes are used to test the algorithm, which shows that it is feasible and can balance the load quite weill.
    • FullText(HTML)
    References(6)
  • DownLoad: Full-Size Img PowerPoint
    DownLoad: Full-Size Img PowerPoint
  • [1]
    Michalakes J, Canfield T, Nanjundiah R, et al. Parallel implementation, validation and performance of MM5, coming of age. Proceedings of the Sixth ECMWF Workshop on the Use of Parallel Processors in Meteorology. World Scientific, River Edge, New Jersey, 1995. 266-276. http://www.gbv.de/dms/tib-ub-hannover/198050437.pdf
    [2]
    Michalakes J, Dudhia J, Gill D, et al. Design of a next-generation regional weather research and forecast model, towards teracomputing. Proceedings of the Eighth ECMWF Workshop on the Use of Parallel Processors in Meteorology. World Scientific, River Edge, New Jersey, 1999. 117-123.
    [3]
    Schattler U. Model development for parallel computers at DWD, making its mark. Proceedings of the Seventh ECMWF Workshop on the Use of parallel Processors in Meteorology. World Scientific, River Edge, New Jersey, 1997. 83-99. http://www.gbv.de/dms/tib-ub-hannover/249689324.pdf
    [4]
    舒继武, 郑纬民, 沈美明, 等, 大规模问题数据并行性能的分析.软件学报, 2000, 11(5):628-633. http://www.cnki.com.cn/Article/CJFDTOTAL-RJXB200005007.htm
    [5]
    Ford R W, Burton P M. Load balancing physics routines, towards teracomputing. Proceedings of the Eighth ECMWF Workshop on the Use of parallel Processors in Meteorology. World Scientific, River Edge, New Jersey, 1999. 147-159. http://www.worldcat.org/oclc/43250502
    [6]
    金之雁, 王鼎兴.一种有限差分格式负载平衡区域分解方法.气象学报, 2002, 60(2):188-193.
    • Relative Articles
    • Supplements(0)
    • Cited By
  • 加载中
  • -->

Catalog

    Figures(2)  / Tables(2)

    Get Citation
    PDF
    XML
    Article views (3514) PDF downloads(1478) Cited by()
    • Received : 2002-02-26
    • Accepted : 2002-06-25
    • Published : 2003-08-31
    Journal Online
    Contact

    © 2020 Journal of Applied Meteorological Science   京ICP备15006967号-1

    Supported by Beijing Renhe Information Technology Co. Ltd

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return