{"created":"2023-05-15T12:36:16.352510+00:00","id":5507,"links":{},"metadata":{"_buckets":{"deposit":"d158489a-2015-4798-b340-d6556c43a124"},"_deposit":{"created_by":3,"id":"5507","owners":[3],"pid":{"revision_id":0,"type":"depid","value":"5507"},"status":"published"},"_oai":{"id":"oai:nitech.repo.nii.ac.jp:00005507","sets":["31"]},"author_link":["9173","19413","8515","8669"],"item_10001_alternative_title_24":{"attribute_name":"その他（別言語等）のタイトル","attribute_value_mlt":[{"subitem_alternative_title":"スクスウ　イタレーション　ノ　イッカツ　サイリヨウ　ニヨル　ヘイレツ　ジゼン　ジッコウ　ノ　コウソクカ"},{"subitem_alternative_title":"A Speed-up Technique for Parallel Early Computation by Collectively Reusing Multiple Iterations"}]},"item_10001_biblio_info_28":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicIssueDates":{"bibliographicIssueDate":"2010-09-17","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"3","bibliographicPageEnd":"43","bibliographicPageStart":"31","bibliographicVolumeNumber":"3","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"}]}]},"item_10001_description_36":{"attribute_name":"内容記述","attribute_value_mlt":[{"subitem_description":"我々は，計算再利用技術に基づく自動メモ化プロセッサ，および，これに値予測に基づく投機マルチスレッド実行を組み合わせた並列事前実行を提案している．従来の並列事前実行機構ではループの各イタレーションを再利用対象の命令区間として抽出していた．本稿では，実行バイナリに変更を加えることなく，複数イタレーションを動的にまとめて再利用対象区間とすることによって，再利用に要するオーバヘッドを削減し，同時に再利用表エントリの効率的な活用を実現する手法を提案する．また，いくつのイタレーションを再利用区間として統合すべきかは対象ループにより異なるため，動的に適切な数を検出するモデルを提案する．SPEC CPU95 FP を用いてシミュレーションにより評価した結果，従来モデルでは最大 40.5%，平均 15.0%であったサイクル数削減率が，最大 57.6%，平均 26.0%まで向上することを確認した．We have proposed an auto-memoization processor based on computation reuse, and merged it with speculative multithreading based on value prediction into a parallel early computation. In the past model, the parallel early computation detects each iterations of loops as reusable blocks. This paper proposes a new parallel early computation model, which integrates plural iterations into a reusable block automatically and dynamically without modifing executable binaries. We also proposes a model for automatically detecting how many iterations should be integrated into one reusable block. Our model reduces the overhead of computation reuse, and further exploits reuse tables. The result of the experiment with SPEC CPU95 FP suite benchmarks shows that proposing method improve the maximum speedup from 40.5% to 57.6%, and the average speedup from 15.0% to 26.0%.","subitem_description_type":"Other"}]},"item_10001_description_38":{"attribute_name":"フォーマット","attribute_value_mlt":[{"subitem_description":"application/pdf","subitem_description_type":"Other"}]},"item_10001_full_name_27":{"attribute_name":"著者別名","attribute_value_mlt":[{"nameIdentifiers":[{},{}],"names":[{"name":"Tsumura, Tomoaki"}]},{"nameIdentifiers":[{},{}],"names":[{"name":"Matsuo, Hiroshi"}]}]},"item_10001_publisher_29":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会"}]},"item_10001_source_id_30":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"03875806","subitem_source_identifier_type":"ISSN"}]},"item_10001_source_id_32":{"attribute_name":"書誌レコードID（NCID）","attribute_value_mlt":[{"subitem_source_identifier":"AA11833852","subitem_source_identifier_type":"NCID"}]},"item_10001_version_type_33":{"attribute_name":"著者版フラグ","attribute_value_mlt":[{"subitem_version_resource":"http://purl.org/coar/version/c_970fb48d4fbd8a85","subitem_version_type":"VoR"}]},"item_creator":{"attribute_name":"著者","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"池谷, 友基"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"津邑, 公暁"},{"creatorName":"ツムラ, トモアキ","creatorNameLang":"ja-Kana"}],"nameIdentifiers":[{},{}]},{"creatorNames":[{"creatorName":"松尾, 啓志"},{"creatorName":"マツオ, ヒロシ","creatorNameLang":"ja-Kana"}],"nameIdentifiers":[{},{}]},{"creatorNames":[{"creatorName":"中島, 康彦"}],"nameIdentifiers":[{}]}]},"item_files":{"attribute_name":"ファイル情報","attribute_type":"file","attribute_value_mlt":[{"accessrole":"open_date","date":[{"dateType":"Available","dateValue":"2017-01-26"}],"displaytype":"detail","filename":"tsumura_2010_acs_1.pdf","filesize":[{"value":"941.7 kB"}],"format":"application/pdf","license_note":"(c)  2010 Information Processing Society of Japan","licensetype":"license_note","mimetype":"application/pdf","url":{"label":"本文_fulltext","url":"https://nitech.repo.nii.ac.jp/record/5507/files/tsumura_2010_acs_1.pdf"},"version_id":"179ecbb7-f2e9-4e00-b533-a492c3b25d56"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourcetype":"journal article","resourceuri":"http://purl.org/coar/resource_type/c_6501"}]},"item_title":"複数イタレーションの一括再利用による並列事前実行の高速化","item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"複数イタレーションの一括再利用による並列事前実行の高速化"}]},"item_type_id":"10001","owner":"3","path":["31"],"pubdate":{"attribute_name":"公開日","attribute_value":"2012-11-07"},"publish_date":"2012-11-07","publish_status":"0","recid":"5507","relation_version_is_last":true,"title":["複数イタレーションの一括再利用による並列事前実行の高速化"],"weko_creator_id":"3","weko_shared_id":3},"updated":"2023-05-15T13:53:37.608268+00:00"}