Implementing MPI in a High - Performance Multithreaded Language MPC++

この論文にアクセスする

この論文をさがす

著者

抄録

We have ported the MPICH implementation of MPI to the high-performance multithreaded programming language MPC++. We discuss our modifications to the design of MPICH to supprt multiple threads. MPICH now runs experimentally on top of MPC++ on a Sun workstation cluster connected by Myrinet and achieves higher performance than standard MPICH on Myrinet TCP/IP on the same hardware.We have ported the MPICH implementation of MPI to the high-performance, multithreaded programming language MPC++. We discuss our modifications to the design of MPICH to supprt multiple threads. MPICH now runs experimentally on top of MPC++ on a Sun workstation cluster connected by Myrinet and achieves higher performance than standard MPICH on Myrinet TCP/IP on the same hardware.

We have ported the MPICH implementation of MPI to the high-performance, multithreaded programming language MPC++. We discuss our modifications to the design of MPICH to support multiple threads. MPICH now runs experimentally on top of MPC++ on a Sun workstation cluster connected by Myrinet and achieves higher performance than standard MPICH on Myrinet TCP/IP on the same hardware.

収録刊行物

  • 情報処理学会研究報告ハイパフォーマンスコンピューティング(HPC)

    情報処理学会研究報告ハイパフォーマンスコンピューティング(HPC) 1996(81(1996-HPC-062)), 141-146, 1996-08-28

    一般社団法人情報処理学会

参考文献:  8件中 1-8件 を表示

各種コード

  • NII論文ID(NAID)
    110002932211
  • NII書誌ID(NCID)
    AN10463942
  • 本文言語コード
    JPN
  • 資料種別
    Technical Report
  • データ提供元
    CJP書誌  NII-ELS  IPSJ 
ページトップへ