• DocumentCode
    1918771
  • Title

    A Non-intrusive Methodology to Improve the Performance of Parallel Applications in High Performance Computing

  • Author

    Luz, Fernando H P ; Taniguchi, Denis ; Sato, Liria M.

  • Author_Institution
    Numerical OffShore Tank Escola Politec., Univ. de Sao Paulo, Sao Paulo, Brazil
  • fYear
    2012
  • fDate
    10-13 Sept. 2012
  • Firstpage
    390
  • Lastpage
    396
  • Abstract
    The effective use of parallel resources is a central topic for high performance computing developers. Another key challenge is the elaboration of smart systems, in which the final user does not need to define the method to distribute the loads in a cluster. In this work, a methodology is proposed to enable the load balance in an automatized and non-intrusive way. A proposal for implementation and achieved speedups are also discussed.
  • Keywords
    parallel programming; resource allocation; high performance computing developers; load balance; non-intrusive methodology; parallel resources; smart systems; Calibration; Computer aided software engineering; High performance computing; Indexes; Load modeling; Program processors; Synchronization; HPC; cluster environment; load balance; scalability;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Parallel Processing Workshops (ICPPW), 2012 41st International Conference on
  • Conference_Location
    Pittsburgh, PA
  • ISSN
    1530-2016
  • Print_ISBN
    978-1-4673-2509-7
  • Type

    conf

  • DOI
    10.1109/ICPPW.2012.56
  • Filename
    6337506
    • Link To Document
    https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=1918771