• DocumentCode
    3588926
  • Title

    Flow Graph Designer: A Tool for Designing and Analyzing Intel® Threading Building Blocks Flow Graphs

  • Author

    Tovinkere, Vasanth ; Voss, Michael

  • fYear
    2014
  • Firstpage
    149
  • Lastpage
    158
  • Abstract
    Flow Graph Designer is a visualization tool that supports the analysis and design of parallel applications that use the Intel® Threading Building Blocks (Intel® TBB) flow graph interface. The flow graph interface allows developers to express the dependency, streaming and data flow graphs present in many domains such as media, gaming, finance, high performance computing and healthcare. Because the flow graph interface introduced a style of programming that is different than loop- and task-parallel approaches, there is a need for new tool support. This paper presents Flow Graph Designer, an experimental graphical tool developed to make the flow graph interface easier to adopt, use, debug and tune. This paper provides a brief introduction to the Intel TBB flow graph interface and an overview of the key features provided by the Flow Graph Designer. Three examples are then presented that demonstrate the utility of this tool for important usage scenarios encountered while designing and analyzing flow graph applications.
  • Keywords
    data flow graphs; data visualisation; multi-threading; software tools; Intel TBB flow graph interface; Intel threading building blocks; data flow graphs; dependency graphs; experimental graphical tool; finance; flow graph applications; flow graph designer; gaming; healthcare; high performance computing; media; parallel applications analysis; parallel applications design; programming style; tool support; visualization tool; Concurrent computing; Engines; Flow graphs; Hardware; Image color analysis; Instruction sets; Parallel processing; code construction; data flow graphs; dependency graphs; performance analysis;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Parallel Processing Workshops (ICCPW), 2014 43rd International Conference on
  • ISSN
    1530-2016
  • Type

    conf

  • DOI
    10.1109/ICPPW.2014.31
  • Filename
    7103449