• 中国精品科技期刊
  • CCF推荐A类中文期刊
  • 计算领域高质量科技期刊T1类
Advanced Search
Li Qiang, Sun Ninghui, Huo Zhigang, Ma Jie. Optimizing MPI Alltoall Communications in Multicore Clusters[J]. Journal of Computer Research and Development, 2013, 50(8): 1744-1754.
Citation: Li Qiang, Sun Ninghui, Huo Zhigang, Ma Jie. Optimizing MPI Alltoall Communications in Multicore Clusters[J]. Journal of Computer Research and Development, 2013, 50(8): 1744-1754.

Optimizing MPI Alltoall Communications in Multicore Clusters

More Information
  • Published Date: August 14, 2013
  • MPI Alltoall is an important collective operation. In multicore clusters, many processes run in a node. On the one hand, shared memory can be adopted to optimize Alltoall communications of small messages by leader-based schemes. However, as these schemes adopt a fixed number of leader processes, the optimal performance can't be obtained for all small messages. On the other hand, processes within a node contend for the same network resource. In Alltoall communications of large messages, many synchronization messages are used. Nevertheless, the contention makes their latency increase many times and the synchronization overhead can't be ingored. To solve these problems, two optimizations are presented. For small messages, the PLP method adopts changeable numbers of leader processes. For large messages, the LSS method reduces the number of synchronization messages from 3N to 2N. The evaluations prove two methods. For small messages, the PLP method always obtains optimal performance. For large messages, the LSS method brings almost constant improvement percentage. The performance is improved by 25% for 32KB and 64KB messages.
  • Related Articles

    [1]Jiang Luyu, Ouyang Dantong, Zhang Qi, Tai Ran, Zhang Liming. Incremental Information Interaction-Based Algorithm for Enumerating MUSes[J]. Journal of Computer Research and Development. DOI: 10.7544/issn1000-1239.202440326
    [2]Bai Tian, Xiao Mingyu. Computational Complexity of Feedback Set and Subset Feedback Set Problems: A Survey[J]. Journal of Computer Research and Development, 2025, 62(1): 104-118. DOI: 10.7544/issn1000-1239.202330693
    [3]Ouyang Dantong, Jia Fengyu, Liu Siguang, Zhang Liming. An Algorithm Based on Extension Rule For Solving #SAT Using Complementary Degree[J]. Journal of Computer Research and Development, 2016, 53(7): 1596-1604. DOI: 10.7544/issn1000-1239.2016.20150032
    [4]Li Shaohua, Feng Qilong, Wang Jianxin, and Chen Jianer. Kernelization for Weighted 3-Set Packing Problem[J]. Journal of Computer Research and Development, 2012, 49(8): 17811-786.
    [5]Qiu Jiangtao, Tang Changjie, Zeng Tao, Liu Yintian. Strategy of Revising Rules for Association Text Classification[J]. Journal of Computer Research and Development, 2009, 46(4): 683-688.
    [6]Zhong Yong, Qin Xiaolin, and Bao Lei. An Association Rule Mining Algorithm of Multidimensional Sets[J]. Journal of Computer Research and Development, 2006, 43(12): 2117-2123.
    [7]Xiong Zhongmin, Hao Zhongxiao. An Approach to Termination Decision for a Rule Set Based on Activation Path and Conditional Formula[J]. Journal of Computer Research and Development, 2006, 43(5): 901-907.
    [8]Hao Zhongxiao, Xiong Zhongmin. An Efficient Algorithm for Computing an Irreducible Rule Set in Active Database[J]. Journal of Computer Research and Development, 2006, 43(2): 281-287.
    [9]Hao Zhongxiao, Ren Chao, Zhao Lingqiang. Termination Analysis of Active Rule Based on Dependency Set[J]. Journal of Computer Research and Development, 2005, 42(12): 2199-2205.
    [10]Tian Daxin, Liu Yanheng, Li Yongli, Tang Yi. A Fast Matching Algorithm and Conflict Detection for Packet Filter Rules[J]. Journal of Computer Research and Development, 2005, 42(7): 1128-1135.

Catalog

    Article views (1153) PDF downloads (609) Cited by()

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return