{"id":29785,"updated":"2025-01-22T17:23:03.932663+00:00","links":{},"created":"2025-01-18T22:59:32.627703+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00029785","sets":["1164:2240:2307:2310"]},"path":["2310"],"owner":"1","recid":"29785","title":["並列数値計算ライブラリの開発"],"pubdate":{"attribute_name":"公開日","attribute_value":"1996-08-28"},"_buckets":{"deposit":"aec53982-ec29-4b50-b874-7e5304f3a456"},"_deposit":{"id":"29785","pid":{"type":"depid","value":"29785","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"並列数値計算ライブラリの開発","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"並列数値計算ライブラリの開発"},{"subitem_title":"Development of Parallel Libraries","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"1996-08-28","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"日本原子力研究所計算科学技術推進センター"},{"subitem_text_value":"日本総合研究所"},{"subitem_text_value":"日本原子力研究所計算科学技術推進センター"},{"subitem_text_value":"日本原子力研究所計算科学技術推進センター"},{"subitem_text_value":"日本原子力研究所計算科学技術推進センター"},{"subitem_text_value":"日本原子力研究所計算科学技術推進センター"},{"subitem_text_value":"日本原子力研究所計算科学技術推進センター"},{"subitem_text_value":"日本原子力研究所計算科学技術推進センター"},{"subitem_text_value":"日本原子力研究所計算科学技術推進センター"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Center for Promotion of Computational Science and Engineering, Japan Atomic Energy Research Institute","subitem_text_language":"en"},{"subitem_text_value":"The Japan Research Institute","subitem_text_language":"en"},{"subitem_text_value":"Center for Promotion of Computational Science and Engineering, Japan Atomic Energy Research Institute","subitem_text_language":"en"},{"subitem_text_value":"Center for Promotion of Computational Science and Engineering, Japan Atomic Energy Research Institute","subitem_text_language":"en"},{"subitem_text_value":"Center for Promotion of Computational Science and Engineering, Japan Atomic Energy Research Institute","subitem_text_language":"en"},{"subitem_text_value":"Center for Promotion of Computational Science and Engineering, Japan Atomic Energy Research Institute","subitem_text_language":"en"},{"subitem_text_value":"Center for Promotion of Computational Science and Engineering, Japan Atomic Energy Research Institute","subitem_text_language":"en"},{"subitem_text_value":"Center for Promotion of Computational Science and Engineering, Japan Atomic Energy Research Institute","subitem_text_language":"en"},{"subitem_text_value":"Center for Promotion of Computational Science and Engineering, Japan Atomic Energy Research Institute","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/29785/files/IPSJ-HPC96062022.pdf"},"date":[{"dateType":"Available","dateValue":"1998-08-28"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-HPC96062022.pdf","filesize":[{"value":"390.2 kB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"14"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"2807c20e-5083-4b3e-bec9-7ca7c14f2e8a","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 1996 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"清水, 大志"},{"creatorName":"佐々木, 誠"},{"creatorName":"市原, 潔"},{"creatorName":"岸田, 則生"},{"creatorName":"鈴木, 惣一朗"},{"creatorName":"佐藤, 滋"},{"creatorName":"田中, 靖久"},{"creatorName":"横川, 三津夫"},{"creatorName":"蕪木, 英雄"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Futoshi, Shimizu","creatorNameLang":"en"},{"creatorName":"Makoto, Sasaki","creatorNameLang":"en"},{"creatorName":"Kiyoshi, Ichihara","creatorNameLang":"en"},{"creatorName":"Norio, Kishida","creatorNameLang":"en"},{"creatorName":"Soichiro, Suzuki","creatorNameLang":"en"},{"creatorName":"Shigeru, Sato","creatorNameLang":"en"},{"creatorName":"Yasuhisa, Tanaka","creatorNameLang":"en"},{"creatorName":"Mitsuo, Yokokawa","creatorNameLang":"en"},{"creatorName":"Hideo, Kaburaki","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10463942","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"近年、演算回路素子等の速度向上が限界となりつつあるために、並列計算は大規模数値シミュレーションの分野において重要な手法となっており、効率及び移植性の良い並列ライプラリが必要とされている。そこで、我々は各種計算機に対応可能であるMPIまたはPVMを用いた分散メモリ型ベクトル並列計算機用数値計算ライブラリの開発を行っている。本論ではHouseholder変換による三重対角化及び2分法を用いた実密対象行列の固有値問題解法ルーチンの開発について報告する。行列のデータは全ての成分を格納する。Householder変換について8プロセッサを使用した並列化による列方向サイクリック方式により分割し、プロセッサ間のデータ転送量を減らすため対称行列の速度向上率は、2000×2000行列に対してParagonで6．0倍である。VPP300では4000×4000行列に対して4．2倍の値を得た。","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"In recent years, since the development of fast processor element has reached the upper limit, parallelism is one of the solutions for massive numerical simulations, and efficient and portable parallel libraries are needed. With MPI or PVM which is not limited to a certain architecture, we are developing parallel subroutine libraries specific for use on parallel vector processors. We report here the development of parallel subroutine library for eigenvalue problem of real symmetric matrix based on the Householder transformation and the bisection method. The matrix is partitioned into columns by the column-wise cyclic decomposition scheme, and all elements of the symmetric matrix are stored in order to reduce data exchanges among processors. For the Householder transformation using eight processors on Paragon, the speedup ratio of 6.0 has been achieved for a matrix of 2000×2000 elements. In the case of the matrix of 4000×4000 elements, the ratio is found to be 4.2 on VPP300.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"134","bibliographic_titles":[{"bibliographic_title":"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）"}],"bibliographicPageStart":"129","bibliographicIssueDates":{"bibliographicIssueDate":"1996-08-28","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"81(1996-HPC-062)","bibliographicVolumeNumber":"1996"}]},"relation_version_is_last":true,"weko_creator_id":"1"}}