Coarse-grain task parallel processing using the openMP backend of the OSCAR multigrain parallelizing compiler

Kazuhisa Ishizaka, Motoki Obata, Hironori Kasahara

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    5 Citations (Scopus)

    Abstract

    This paper describes automatic coarse grain parallel process-ing on a shared memory multiprocessor system using a newly developed OpenMP backend of OSCAR multigrain parallelizing compiler for from single chip multiprocessor to a high performance multiprocessor and a heterogeneous supercomputer cluster. OSCAR multigrain parallelizing compiler exploits coarse grain task parallelism and near fine grain paral-lelism in addition to traditional loop parallelism. The OpenMP backend generates parallelized Fortran code with OpenMP directives based on an-alyzed multigrain parallelism by middle path of OSCAR compiler from an ordinary Fortran source program. The performance of multigrain par-allel processing function by OpenMP backend is evaluated on an off the shelf eight processor SMP machine, IBM RS6000. The evaluation shows that the multigrain parallel processing gives us more than 2 times speed up compared with a commercial loop parallelizing compiler, IBM XL Fortran compiler, on the SMP machine.

    Original languageEnglish
    Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    PublisherSpringer Verlag
    Pages457-470
    Number of pages14
    Volume1940
    ISBN (Print)9783540411284
    Publication statusPublished - 2000
    Event3rd International Symposium on High Performance Computing, ISHPC 2000 - Tokyo, Japan
    Duration: 2000 Oct 162000 Oct 18

    Publication series

    NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume1940
    ISSN (Print)03029743
    ISSN (Electronic)16113349

    Other

    Other3rd International Symposium on High Performance Computing, ISHPC 2000
    CountryJapan
    CityTokyo
    Period00/10/1600/10/18

    Fingerprint

    Parallelizing Compilers
    OpenMP
    Parallel Processing
    Parallelism
    Processing
    Compiler
    Supercomputers
    Shared-memory multiprocessors
    Chip multiprocessors
    Multiprocessor Systems
    Supercomputer
    Multiprocessor
    Data storage equipment
    Speedup
    High Performance
    Path
    Evaluation

    ASJC Scopus subject areas

    • Computer Science(all)
    • Theoretical Computer Science

    Cite this

    Ishizaka, K., Obata, M., & Kasahara, H. (2000). Coarse-grain task parallel processing using the openMP backend of the OSCAR multigrain parallelizing compiler. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1940, pp. 457-470). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1940). Springer Verlag.

    Coarse-grain task parallel processing using the openMP backend of the OSCAR multigrain parallelizing compiler. / Ishizaka, Kazuhisa; Obata, Motoki; Kasahara, Hironori.

    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1940 Springer Verlag, 2000. p. 457-470 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1940).

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Ishizaka, K, Obata, M & Kasahara, H 2000, Coarse-grain task parallel processing using the openMP backend of the OSCAR multigrain parallelizing compiler. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 1940, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1940, Springer Verlag, pp. 457-470, 3rd International Symposium on High Performance Computing, ISHPC 2000, Tokyo, Japan, 00/10/16.
    Ishizaka K, Obata M, Kasahara H. Coarse-grain task parallel processing using the openMP backend of the OSCAR multigrain parallelizing compiler. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1940. Springer Verlag. 2000. p. 457-470. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    Ishizaka, Kazuhisa ; Obata, Motoki ; Kasahara, Hironori. / Coarse-grain task parallel processing using the openMP backend of the OSCAR multigrain parallelizing compiler. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1940 Springer Verlag, 2000. pp. 457-470 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
    @inproceedings{66a6b718b33d4819a4e0688a20c280e5,
    title = "Coarse-grain task parallel processing using the openMP backend of the OSCAR multigrain parallelizing compiler",
    abstract = "This paper describes automatic coarse grain parallel process-ing on a shared memory multiprocessor system using a newly developed OpenMP backend of OSCAR multigrain parallelizing compiler for from single chip multiprocessor to a high performance multiprocessor and a heterogeneous supercomputer cluster. OSCAR multigrain parallelizing compiler exploits coarse grain task parallelism and near fine grain paral-lelism in addition to traditional loop parallelism. The OpenMP backend generates parallelized Fortran code with OpenMP directives based on an-alyzed multigrain parallelism by middle path of OSCAR compiler from an ordinary Fortran source program. The performance of multigrain par-allel processing function by OpenMP backend is evaluated on an off the shelf eight processor SMP machine, IBM RS6000. The evaluation shows that the multigrain parallel processing gives us more than 2 times speed up compared with a commercial loop parallelizing compiler, IBM XL Fortran compiler, on the SMP machine.",
    author = "Kazuhisa Ishizaka and Motoki Obata and Hironori Kasahara",
    year = "2000",
    language = "English",
    isbn = "9783540411284",
    volume = "1940",
    series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
    publisher = "Springer Verlag",
    pages = "457--470",
    booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

    }

    TY - GEN

    T1 - Coarse-grain task parallel processing using the openMP backend of the OSCAR multigrain parallelizing compiler

    AU - Ishizaka, Kazuhisa

    AU - Obata, Motoki

    AU - Kasahara, Hironori

    PY - 2000

    Y1 - 2000

    N2 - This paper describes automatic coarse grain parallel process-ing on a shared memory multiprocessor system using a newly developed OpenMP backend of OSCAR multigrain parallelizing compiler for from single chip multiprocessor to a high performance multiprocessor and a heterogeneous supercomputer cluster. OSCAR multigrain parallelizing compiler exploits coarse grain task parallelism and near fine grain paral-lelism in addition to traditional loop parallelism. The OpenMP backend generates parallelized Fortran code with OpenMP directives based on an-alyzed multigrain parallelism by middle path of OSCAR compiler from an ordinary Fortran source program. The performance of multigrain par-allel processing function by OpenMP backend is evaluated on an off the shelf eight processor SMP machine, IBM RS6000. The evaluation shows that the multigrain parallel processing gives us more than 2 times speed up compared with a commercial loop parallelizing compiler, IBM XL Fortran compiler, on the SMP machine.

    AB - This paper describes automatic coarse grain parallel process-ing on a shared memory multiprocessor system using a newly developed OpenMP backend of OSCAR multigrain parallelizing compiler for from single chip multiprocessor to a high performance multiprocessor and a heterogeneous supercomputer cluster. OSCAR multigrain parallelizing compiler exploits coarse grain task parallelism and near fine grain paral-lelism in addition to traditional loop parallelism. The OpenMP backend generates parallelized Fortran code with OpenMP directives based on an-alyzed multigrain parallelism by middle path of OSCAR compiler from an ordinary Fortran source program. The performance of multigrain par-allel processing function by OpenMP backend is evaluated on an off the shelf eight processor SMP machine, IBM RS6000. The evaluation shows that the multigrain parallel processing gives us more than 2 times speed up compared with a commercial loop parallelizing compiler, IBM XL Fortran compiler, on the SMP machine.

    UR - http://www.scopus.com/inward/record.url?scp=77954410221&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=77954410221&partnerID=8YFLogxK

    M3 - Conference contribution

    AN - SCOPUS:77954410221

    SN - 9783540411284

    VL - 1940

    T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    SP - 457

    EP - 470

    BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    PB - Springer Verlag

    ER -