{"updated":"2025-01-20T09:00:34.919619+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00169435","sets":["934:1119:8503:8746"]},"path":["8746"],"owner":"11","recid":"169435","title":["メニーコアにおける派生データ型を用いたMPIノード内通信の高速化"],"pubdate":{"attribute_name":"公開日","attribute_value":"2016-07-14"},"_buckets":{"deposit":"86065d45-0329-4efb-a592-81fce7b4d71c"},"_deposit":{"id":"169435","pid":{"type":"depid","value":"169435","revision_id":0},"owners":[11],"status":"published","created_by":11},"item_title":"メニーコアにおける派生データ型を用いたMPIノード内通信の高速化","author_link":["340434","340433","340435","340428","340429","340427","340431","340432","340430","340426"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"メニーコアにおける派生データ型を用いたMPIノード内通信の高速化"},{"subitem_title":"Accelerating MPI Intra-node Communication Using Derived Data Types on Many-core","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"[最適化] メニーコア，OS，MPI，ノード内通信，派生データ型","subitem_subject_scheme":"Other"}]},"item_type_id":"3","publish_date":"2016-07-14","item_3_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"株式会社日立製作所研究開発グループ／慶應義塾大学"},{"subitem_text_value":"株式会社日立製作所研究開発グループ"},{"subitem_text_value":"理化学研究所計算科学研究機構"},{"subitem_text_value":"理化学研究所計算科学研究機構"},{"subitem_text_value":"慶應義塾大学"}]},"item_3_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Hitachi, Ltd., Research & Development Group / KEIO University","subitem_text_language":"en"},{"subitem_text_value":"Hitachi, Ltd., Research & Development Group","subitem_text_language":"en"},{"subitem_text_value":"RIKEN AICS","subitem_text_language":"en"},{"subitem_text_value":"RIKEN AICS","subitem_text_language":"en"},{"subitem_text_value":"KEIO University","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/169435/files/IPSJ-TACS0902005.pdf","label":"IPSJ-TACS0902005.pdf"},"date":[{"dateType":"Available","dateValue":"2018-07-14"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-TACS0902005.pdf","filesize":[{"value":"2.0 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"16"},{"tax":["include_tax"],"price":"0","billingrole":"11"},{"tax":["include_tax"],"price":"0","billingrole":"14"},{"tax":["include_tax"],"price":"0","billingrole":"15"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"a0b1f0c1-6e56-452d-bc0c-12a9cf7649dc","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2016 by the Information Processing Society of Japan"}]},"item_3_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"島田, 明男"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"須藤, 敦之"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"堀, 敦史"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"石川, 裕"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"河野, 健二"}],"nameIdentifiers":[{}]}]},"item_3_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Akio, Shimada","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Atsushi, Sutoh","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Atsushi, Hori","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Yutaka, Ishikawa","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kenji, Kono","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_3_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11833852","subitem_source_identifier_type":"NCID"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_6501","resourcetype":"journal article"}]},"item_3_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"1882-7829","subitem_source_identifier_type":"ISSN"}]},"item_3_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"近年，マルチコアプロセッサのコア数は増加する傾向にある．また，低性能だが電力効率が高いコアを多数集約して高い並列処理性能を実現するメニーコアプロセッサの活用も進んでおり，HPCシステムを構成するノード1台あたりのコア数は飛躍的に増加してきている．Partitioned Virtual Address Space（PVAS）は，このようなメニーコア環境において，効率的に並列処理を実行するための新たなタスクモデルである．PVASを用いると，並列処理を行うノード内のプロセス群を同一アドレス空間で動作させることが可能になる．よって，アドレス空間の境界越しにデータを送受信するためのオーバヘッドなしに，ノード内通信を実行することができる．本研究では，PVASを派生データ型を用いるMessage Passing Interface（MPI）ノード内通信に適用し，同一ノードのプロセスがメモリ上で不連続なデータを高速に送受信することを可能にした．ベンチマークプログラムによって高速化したMPIノード内通信を評価したところ，通信遅延を低減させ，実行性能を最大で約21%向上させることができた．","subitem_description_type":"Other"}]},"item_3_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Recently, the number of cores in multi-core processors is expected to continue to grow and many-core processors, which aggregate massive number of weak but power-efficient cores, are widely used. Then, the number of cores in an HPC system node grows rapidly. Partitioned Virtual Address Space (PVAS) is a new task model, which enables efficient parallel processing in such many-core systems. The PVAS task model allows multiple processes to run in the same address space, which means that the processes running on the PVAS task model can exchange data without overheads for crossing address space boundaries. In this paper, we apply the PVAS task model to the Message Passing Interface (MPI) intra-node communication using Derived Data Types and accelerate the transfer of non-contiguous data. The accelerated MPI intra-node communication reduces communication latency and improves application performance by up to 21%.","subitem_description_type":"Other"}]},"item_3_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"63","bibliographic_titles":[{"bibliographic_title":"情報処理学会論文誌コンピューティングシステム（ACS）"}],"bibliographicPageStart":"46","bibliographicIssueDates":{"bibliographicIssueDate":"2016-07-14","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"2","bibliographicVolumeNumber":"9"}]},"relation_version_is_last":true,"weko_creator_id":"11"},"created":"2025-01-19T00:40:12.719112+00:00","id":169435,"links":{}}