• DocumentCode
    2323127
  • Title

    Improving the Task Stealing in Intel Threading Building Blocks

  • Author

    Lu, Shiyi ; Li, Qing

  • Author_Institution
    Dept. of Comput. Sci., ShangHai Univ., Shanghai, China
  • fYear
    2011
  • fDate
    10-12 Oct. 2011
  • Firstpage
    343
  • Lastpage
    346
  • Abstract
    The Intel Threading Building Blocks (TBB)is a run-time library[1] for parallel programming based on C++. The TBB programming environment facilitates programmers to express concurrency in terms of parallel tasks rather than parallel threads. TBB provide programmers with parallel algorithms and concurrent containers, thus programmers can parallel a program without considering how to map the tasks on to the physical threads. In fact, all the paralleling task work is based on the task scheduler. When stealing a task, the task scheduler select the victim thread randomly. In this paper, we first study and characterize the TBB´s random stealing. Then we provide two method to predicate the threads´ criticality, using the methods to improve the performance of TBB´s task stealing. We did some experiments to validate the achievement in improving the performance of TBB´s task stealing.
  • Keywords
    C++ language; parallel algorithms; parallel programming; C++; Intel threading building blocks; TBB programming environment; parallel algorithms; parallel programming; paralleling task work; run-time library; task scheduler; task stealing; Benchmark testing; Computer architecture; Concurrent computing; Dynamic scheduling; Instruction sets; Parallel processing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Cyber-Enabled Distributed Computing and Knowledge Discovery (CyberC), 2011 International Conference on
  • Conference_Location
    Beijing
  • Print_ISBN
    978-1-4577-1827-4
  • Type

    conf

  • DOI
    10.1109/CyberC.2011.61
  • Filename
    6079453