• DocumentCode
    2959917
  • Title

    ScalaBenchGen: Auto-Generation of Communication Benchmarks Traces

  • Author

    Wu, Xing ; Deshpande, Vivek ; Mueller, Frank

  • Author_Institution
    Dept. of Comput. Sci., North Carolina State Univ., Raleigh, NC, USA
  • fYear
    2012
  • fDate
    21-25 May 2012
  • Firstpage
    1250
  • Lastpage
    1260
  • Abstract
    Benchmarks are essential for evaluating HPC hardware and software for petascale machines and beyond. But benchmark creation is a tedious manual process. As a result, benchmarks tend to lag behind the development of complex scientific codes. This work contributes an automated approach to the creation of communication benchmarks. Given an MPI application, we utilize Scala Trace, a loss less and scalable framework to trace communication operations and execution time while abstracting away the computations. A single trace file that reflects the behavior of all nodes is subsequently expanded to C source code by a novel code generator. This resulting benchmark code is compact, portable, human-readable, and accurately reflects the original application´s communication characteristics and runtime characteristics. Experimental results demonstrate that generated source code of benchmarks preserves both the communication patterns and the wall clock-time behavior of the original application. Such automatically generated benchmarks not only shorten the transition from application development to benchmark extraction but also facilitate code obfuscation, which is essential for benchmark extraction from commercial and restricted applications.
  • Keywords
    application program interfaces; benchmark testing; message passing; program compilers; program diagnostics; software performance evaluation; C source code; HPC hardware evaluation; HPC software evaluation; MPI application; ScalaBenchGen; ScalaTrace; benchmark extraction; code generator; code obfuscation; communication benchmark creation; communication benchmarks trace auto-generation; message passing interface; petascale machines; trace communication operations; trace execution time; wall clock-time behavior; Distributed processing; Benchmark Generation; Performance; ScalaTrace;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Parallel & Distributed Processing Symposium (IPDPS), 2012 IEEE 26th International
  • Conference_Location
    Shanghai
  • ISSN
    1530-2075
  • Print_ISBN
    978-1-4673-0975-2
  • Type

    conf

  • DOI
    10.1109/IPDPS.2012.114
  • Filename
    6267927