{"links":{},"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00145098","sets":["1164:2240:7894:8332"]},"path":["8332"],"owner":"11","recid":"145098","title":["京・FX10における倍々精度演算の高速化"],"pubdate":{"attribute_name":"公開日","attribute_value":"2015-09-23"},"_buckets":{"deposit":"870dfad7-fe15-4655-9c11-abd7c0d595d7"},"_deposit":{"id":"145098","pid":{"type":"depid","value":"145098","revision_id":0},"owners":[11],"status":"published","created_by":11},"item_title":"京・FX10における倍々精度演算の高速化","author_link":["222122","222120","222117","222119","222121","222118"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"京・FX10における倍々精度演算の高速化"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"数値計算","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2015-09-23","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"工学院大学"},{"subitem_text_value":"筑波大学"},{"subitem_text_value":"工学院大学"},{"subitem_text_value":"工学院大学"},{"subitem_text_value":"国立研究開発法人理化学研究所計算科学研究機構"},{"subitem_text_value":"国立研究開発法人理化学研究所計算科学研究機構"}]},"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/145098/files/IPSJ-HPC15151015.pdf","label":"IPSJ-HPC15151015.pdf"},"date":[{"dateType":"Available","dateValue":"2017-09-23"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-HPC15151015.pdf","filesize":[{"value":"817.9 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":"680ce8a7-9dc6-43f1-bd8d-b895c587970e","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2015 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"佐々木, 信一"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"菱沼, 利彰"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"藤井, 昭宏"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"田中, 輝雄"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"椋木, 大地"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"今村, 俊幸"}],"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_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8841","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"倍々精度演算は Bailey らが考案した ”Double-Double” 精度のアルゴリズム [1] を用いて，倍精度演算の組合せにより四倍精度演算を行う手法である．倍々精度乗算は FMA （積和） 演算器を搭載するプロセッサでは積和演算の中間結果を高精度で保持できアルゴリズム的に演算量を削減できる （FMA アルゴリズム）．本研究ではスーパコンピュータ京 [6] および FX10 に搭載されている FMA 演算器と SIMD（Single Instruction Multiple Data） 演算機構を用いて，これらに最適なベクトル演算および疎行列ベクトル積 （疎行列部は倍精度） の高速化手法を提案する．実験の結果，ベクトル演算ではキャッシュ容量に収まる範囲では本手法適用前に比べ約 5 倍，収まらない範囲ではメモリ性能に制約を受けるまで高速化できること，疎行列ベクトル積では高速化手法によりフロリダ行列に対しては最大 4.5 倍の高速化したことを確認した．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"7","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":"2015-09-23","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"15","bibliographicVolumeNumber":"2015-HPC-151"}]},"relation_version_is_last":true,"weko_creator_id":"11"},"updated":"2025-01-20T18:29:27.830799+00:00","created":"2025-01-19T00:20:46.784199+00:00","id":145098}