{"id":234568,"created":"2025-01-19T01:36:21.975888+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00234568","sets":["1164:1579:11464:11617"]},"path":["11617"],"owner":"44499","recid":"234568","title":["GPUにおける距離指定型命令セットを用いた軽量なアウトオブオーダー実行"],"pubdate":{"attribute_name":"公開日","attribute_value":"2024-06-03"},"_buckets":{"deposit":"48d4e380-9e65-4303-93de-91c14bea593b"},"_deposit":{"id":"234568","pid":{"type":"depid","value":"234568","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"GPUにおける距離指定型命令セットを用いた軽量なアウトオブオーダー実行","author_link":["639342","639341","639343","639344","639340"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"GPUにおける距離指定型命令セットを用いた軽量なアウトオブオーダー実行"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"GPU・DPU","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2024-06-03","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"東京大学"},{"subitem_text_value":"名古屋工業大学"},{"subitem_text_value":"東京大学"},{"subitem_text_value":"東京大学"},{"subitem_text_value":"東京大学"}]},"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/234568/files/IPSJ-ARC24257024.pdf","label":"IPSJ-ARC24257024.pdf"},"date":[{"dateType":"Available","dateValue":"2026-06-03"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-ARC24257024.pdf","filesize":[{"value":"2.3 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":"44"}],"accessrole":"open_date","version_id":"2303cc4f-58f4-45d8-9fd6-967fd4c48630","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2024 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"松尾, 玲央馬"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"小泉, 透"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"入江, 英嗣"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"坂井, 修一"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"塩谷, 亮太"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10096105","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_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8574","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"Graphic Processing Unit（GPU）はデータ並列性を利用することで高いスループットを実現するプロセッサである．その高いスループットを実現する機能の一つがマルチスレッド実行であり，一般に GPU は多くのスレッドを同時に実行することで様々なレイテンシを隠蔽する．そのため，スレッド内で遅延を隠蔽する高コストな out-of-order 実行は，GPU では一般に行われない．しかし，GPU のワークロードにおいても多くの命令レベル並列性が含まれており，out-of-order 実行によってこれを抽出することでさらに GPU の性能を向上させることができる．本研究では，GPU において非常に低コストな out-of-order 実行を実現する TURBULENCE アーキテクチャを提案する．TURBULENCE は，レジスタ番号ではなく命令間距離によってオペランドを指定する表現を導入した新しい ISA と，その新しい ISA を実行するマイクロアーキテクチャからなる．この距離に基づくオペランド表現を用いた命令では，命令間における偽の依存が生じない性質がある．この性質を利用し，本研究ではリネーム・ロジックやロード・ストア・キューなどの高コストな回路を導入せずに，GPU においてコスト効率の良い out-of-order 実行を実現する．シミュレーションによる評価の結果，TURBULENCEは ベースライン構成の GPU と比較して消費エネルギーを増加させることなく 23.7% 性能を向上させることができた．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"8","bibliographic_titles":[{"bibliographic_title":"研究報告システム・アーキテクチャ（ARC）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2024-06-03","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"24","bibliographicVolumeNumber":"2024-ARC-257"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"updated":"2025-01-19T09:46:02.475160+00:00","links":{}}