{"id":66883,"created":"2025-01-18T23:27:31.035424+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00066883","sets":["1164:2240:5632:5925"]},"path":["5925"],"owner":"10","recid":"66883","title":["GPUによる4倍精度BLASの実装と評価"],"pubdate":{"attribute_name":"公開日","attribute_value":"2009-11-23"},"_buckets":{"deposit":"16b02f6f-4236-4a0e-8176-b0db891c9c7e"},"_deposit":{"id":"66883","pid":{"type":"depid","value":"66883","revision_id":0},"owners":[10],"status":"published","created_by":10},"item_title":"GPUによる4倍精度BLASの実装と評価","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"GPUによる4倍精度BLASの実装と評価"},{"subitem_title":"Implementation and Evaluation of Quadruple Precision BLAS on GPU","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"GPUによる数値計算","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2009-11-23","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"筑波大学システム情報工学研究科"},{"subitem_text_value":"筑波大学システム情報工学研究科"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Graduate School of Systems and Information Engineering, University of Tsukuba","subitem_text_language":"en"},{"subitem_text_value":"Graduate School of Systems and Information Engineering, University of Tsukuba","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/66883/files/IPSJ-HPC09123013.pdf"},"date":[{"dateType":"Available","dateValue":"2011-11-23"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-HPC09123013.pdf","filesize":[{"value":"354.1 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":"e0850250-1fc5-426f-bdd5-73996e5c8483","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2009 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"椋木, 大地"},{"creatorName":"高橋, 大介"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Daichi, Mukunoki","creatorNameLang":"en"},{"creatorName":"Daisuke, Takahashi","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":"本稿では GPU （Graphics Processing Unit） で動作する 4 倍精度 BLAS （Basic Linear Algebra Subprograms） の実装を行った．GPU はメモリ性能に対して演算性能が高く，性能を引き出すには演算密度の高いアプリケーションが求められる．我々は演算密度の高い処理として倍精度演算を組み合わせた Double–Double 型 （DD 型） の 4 倍精度演算と行列積に着目し，GPU 向けの汎用計算開発環境である CUDA （Compute Unified Device Architecture） を用いて，DD 型 4 倍精度 BLAS を実装した．NVIDIA Tesla C1060 における性能評価では，Intel Core i7 920 上で実行した CPU 向けの DD 型 4 倍精度 BLAS である MBLAS と比較し，DD 型 4 倍精度行列積 （DDGEMM） で最大約 30 倍の性能を得た．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"We implemented a quadruple precision BLAS (Basic Linear Algebra Subprograms) on GPU (Graphics Processing Unit). Since GPU computing performance is much higher than memory bandwidth, it needs the computationintensive applications to give its best performance. As a computation-intensive operation, we focused on Double–Double (DD) quadruple precision operations combined double precision operations and matrix multiplication. We implemented a quadruple precision BLAS using CUDA (Compute Unified Device Architecture) which is general purpose computing environment for GPUs. The experimental results on NVIDIA Tesla C1060 show that DD precision matrix multiplication (DDGEMM) runs maximum 30 times faster than MBLAS on Intel Core i7 920.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告ハイパフォーマンスコンピューティング（HPC）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2009-11-23","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"13","bibliographicVolumeNumber":"2009-HPC-123"}]},"relation_version_is_last":true,"weko_creator_id":"10"},"updated":"2025-01-22T00:50:40.110619+00:00","links":{}}