Hierarchical network coding for collective communication on HPC interconnects

Ahmed Shalaby, Mohamed El Sayed Ragab, Victor Goulart, Ikki Fujiwara, Michihiro Koibuchi

    Research output: Contribution to conferencePaperpeer-review

    1 Citation (Scopus)

    Abstract

    Network bandwidth is a performance concern especially for collective communication because the bisection bandwidth of recent supercomputers is far less than their full bisection bandwidth. In this context we propose to exploit the use of a network coding technique to reduce the number of unicasts and the size of transferred data generated by latency-sensitive collective communication in supercomputers. Our proposed network coding scheme has a hierarchical multicasting structure with intra-group and inter-group unicasts. Quantitative analysis show that the aggregate path hop counts by our hierarchical network coding decrease as much as 94% when compared to conventional unicast-based multicasts. We validate these results by cycle-accurate network simulations. In 1,024-switch networks, the network reduces the execution time of collective communication as much as 64%. We also show that our hierarchical network coding is beneficial for any packet size.

    Original languageEnglish
    Pages98-102
    Number of pages5
    DOIs
    Publication statusPublished - 2014
    Event2014 22nd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing, PDP 2014 - Turin, Italy
    Duration: Feb 12 2014Feb 14 2014

    Other

    Other2014 22nd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing, PDP 2014
    Country/TerritoryItaly
    CityTurin
    Period2/12/142/14/14

    All Science Journal Classification (ASJC) codes

    • Computer Networks and Communications
    • Software

    Fingerprint

    Dive into the research topics of 'Hierarchical network coding for collective communication on HPC interconnects'. Together they form a unique fingerprint.

    Cite this