• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Development of Innovative Auto-tuning Middleware With Deep Learning

Research Project

Project/Area Number 18K19782
Research Category

Grant-in-Aid for Challenging Research (Exploratory)

Allocation TypeMulti-year Fund
Review Section Medium-sized Section 60:Information science, computer engineering, and related fields
Research InstitutionNagoya University

Principal Investigator

Katagiri Takahiro  名古屋大学, 情報基盤センター, 教授 (40345434)

Co-Investigator(Kenkyū-buntansha) 大島 聡史  名古屋大学, 情報基盤センター, 准教授 (40570081)
田中 輝雄  工学院大学, 情報学部(情報工学部), 教授 (90622837)
Project Period (FY) 2018-06-29 – 2020-03-31
Project Status Completed (Fiscal Year 2019)
Budget Amount *help
¥6,240,000 (Direct Cost: ¥4,800,000、Indirect Cost: ¥1,440,000)
Fiscal Year 2019: ¥3,120,000 (Direct Cost: ¥2,400,000、Indirect Cost: ¥720,000)
Fiscal Year 2018: ¥3,120,000 (Direct Cost: ¥2,400,000、Indirect Cost: ¥720,000)
Keywordsディープラーニング / 自動チューニング / 前処理方式選択 / 疎行列反復解法 / Xabclib / GpGPU / FIBER方式 / 実行時最適化 / 数値計算ライブラリ / OpenATLib
Outline of Final Research Achievements

Dramatically increase of difficulties for tuning work caused by advanced computer architectures, such as GPUs and many-core CPUs with ability of approximately 300 threads execution, requires a method to obtain maximum performance of software without manual tuning works, which is known as auto-tuning (AT) technology. On the other hand, technology of Deep Learning (DL) is dramatically progressing in several fields. Although DL is one of AT methods, there is little study for AT method with DL. Hence in this research, we have studied as follows: (1) Tuning performance parameters for numerical libraries; (2) Developing an AT basic interface; (3) Adaptation of super-computers.

Academic Significance and Societal Importance of the Research Achievements

(1)数値計算ライブラリにおいて収束性に影響し実行時間に大きな影響を及ぼす前処理選択がある。本研究では前処理選択へ活用できるDLを用いたAT方式を開発した。これにより、数値計算を低いコストで高性能実行できる環境に貢献し、ものつくり等の生産性の向上に資する。(2)提案するAT方式の実用化に向け実行時の性能の揺らぎに対しても追随できるように改良を行なったことで、より堅牢なATシステムの実現に資する。(3)GPUやメニーコア環境における数値計算コードの最適化行うことで、最新計算機環境における最適化と性能評価のためのコードやデータを集め、高性能数値計算プログラム開発のコスト削減に資する。

Report

(3 results)
  • 2019 Annual Research Report   Final Research Report ( PDF )
  • 2018 Research-status Report
  • Research Products

    (32 results)

All 2020 2019 2018 Other

All Int'l Joint Research (2 results) Journal Article (6 results) (of which Int'l Joint Research: 1 results,  Peer Reviewed: 3 results) Presentation (22 results) (of which Int'l Joint Research: 11 results,  Invited: 3 results) Remarks (1 results) Funded Workshop (1 results)

  • [Int'l Joint Research] 國家理論科學研究中心(台湾)/国立中央大学(台湾)(その他の国・地域)

    • Related Report
      2019 Annual Research Report
  • [Int'l Joint Research] ローレンスバークレー国立研究所(米国)

    • Related Report
      2019 Annual Research Report
  • [Journal Article] Performance evaluation of the MODYLAS application on modern multi-core and many-core environments2019

    • Author(s)
      Satoshi Ohshima, Soichiro Suzuki, Tatsuya Sakashita, Masao Ogino, Takahiro Katagiri, Yoshimichi Andoh
    • Journal Title

      2019 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)

      Volume: - Pages: 787-796

    • DOI

      10.1109/ipdpsw.2019.00129

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Optimization of Numerous Small Dense-Matrix-Vector Multiplications in H-matrix Arithmetic on GPU2019

    • Author(s)
      Satoshi Ohshima, Ichitaro Yamazaki, Akihiro Ida, Rio Yokota
    • Journal Title

      Auto-Tuning for Multicore and GPU (ATMG) In conjunction with the IEEE MCSoC-19

      Volume: なし Pages: 9-16

    • DOI

      10.1109/mcsoc.2019.00009

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] 性能パラメータ推定における評価対象プログラムの実行時間の揺らぎに対応した自動チューニング手法の提案2019

    • Author(s)
      関直人,范谷瑛,多部田敏樹,藤井昭宏, 田中輝雄
    • Journal Title

      研究報告ハイパフォーマンスコンピューティング(HPC)

      Volume: 2019-HPC-169 Pages: 1-8

    • Related Report
      2019 Annual Research Report
  • [Journal Article] Performance evaluation of the MODYLAS application on modern multi-core and many-core environments2019

    • Author(s)
      Satoshi Ohshima, Soichiro Suzuki, Tatsuya Sakashita, Masao Ogino, Takahiro Katagiri, Yoshimichi Andoh
    • Journal Title

      Proc. of IEEE IPDPSW2019

      Volume: - Pages: 1-10

    • Related Report
      2018 Research-status Report
    • Peer Reviewed / Int'l Joint Research
  • [Journal Article] ソフトウェア自動チューニングにおける反復2次元d-Spline探索法の提案と評価2019

    • Author(s)
      范 谷瑛,関 直人,多部田 敏樹,藤井 昭宏,田中 輝雄
    • Journal Title

      研究報告ハイパフォーマンスコンピューティング(HPC)

      Volume: 2019-HPC-168 Pages: 1-8

    • Related Report
      2018 Research-status Report
  • [Journal Article] 512bit SIMD環境における分子動力学アプリケーションMODYLASの性能評価2018

    • Author(s)
      大島聡史, 鈴木惣一朗, 坂下逹哉, 荻野正雄, 片桐孝洋, 安藤嘉倫
    • Journal Title

      研究報告ハイパフォーマンスコンピューティング(HPC)

      Volume: 2018-HPC-166 Pages: 1-9

    • Related Report
      2018 Research-status Report
  • [Presentation] Autotuning by Changing Directives and Number of Threads in OpenMP using ppOpen-AT2020

    • Author(s)
      Toma Sakurai, Takahiro Katagiri, Satoshi Ohshima, Toru Nagai
    • Organizer
      International Conference on High Performance Computing in Asia-Pacific Region (HPCAsia2020)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Towards Auto-tuning Technology in Exascale Era2020

    • Author(s)
      Takahiro Katagiri
    • Organizer
      CANDAR'19 (The Seventh International Symposium on Computing and Networking)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research / Invited
  • [Presentation] OpenACCを用いたGKVベンチマークの並列化2020

    • Author(s)
      森下誠,大島聡史,片桐孝洋,永井亨
    • Organizer
      情報処理学会第82回全国大会
    • Related Report
      2019 Annual Research Report
  • [Presentation] 外乱のある環境での分散深層学習の性能評価2020

    • Author(s)
      山梨祥平,大島聡史,永井亨,片桐孝洋
    • Organizer
      情報処理学会第82回全国大会
    • Related Report
      2019 Annual Research Report
  • [Presentation] 医用画像処理におけるLDDMMのGPU高速化2020

    • Author(s)
      杉浦拓未,大島聡史,中島大地,片桐孝洋,横田達也,本谷秀堅,永井亨
    • Organizer
      情報処理学会第82回全国大会
    • Related Report
      2019 Annual Research Report
  • [Presentation] Windows MR+Unityの環境におけるプラズマ乱流シミュレーションの可視化2020

    • Author(s)
      北澤修太, 沼波政倫, 大谷寛明, 片桐孝洋, 大島聡史, 永井亨
    • Organizer
      先進的可視化環境を用いた可視化情報の研究会(VR2019)
    • Related Report
      2019 Annual Research Report
  • [Presentation] Stable Automatic Tuning Method for Performance Fluctuation2020

    • Author(s)
      Naoto seki, Toshiki Tabeta, Akihiro Fujii, Teruo Tanaka
    • Organizer
      2020 SIAM Conference on Parallel Processing for Scientific Computing (PP20)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] An Optimization technology of Software Auto-Tuning Applied to Machine Learning Software2020

    • Author(s)
      Toshiki Tabeta, Naoto Seki, Akihiro Fujii, Teruo Tanaka, Hiroyuki Takizawa
    • Organizer
      International Conference on High Performance Computing in Asia-Pacific Region (HPCAsia2020)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Trying to accelerate many small BLAS calculations on GPU2019

    • Author(s)
      Satoshi Ohshima
    • Organizer
      ATAT in HPSC (2019 Conference on Advanced Topics and Auto Tuning in High-Performance Scientific Computing)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research / Invited
  • [Presentation] Performance Evaluation of Accurate Matrix-matrix Multiplications on GPU Using Sparse Matrix Multiplications2019

    • Author(s)
      Fumiya Ishiguro, Takahiro Katagiri, Satoshi Ohshima, Toru Nagai
    • Organizer
      International Conference on High Performance Computing in Asia-Pacific Region (HPCAsia2020)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] ポストムーア時代に向けた自動チューニングに向けて ~スレッド数の動的最適化~2019

    • Author(s)
      片桐孝洋、櫻井刀麻
    • Organizer
      第24回計算工学講演会
    • Related Report
      2019 Annual Research Report
  • [Presentation] ポストムーア時代の計算機環境における数値計算カーネル実装の変革2019

    • Author(s)
      片桐孝洋
    • Organizer
      2019年並列/分散/協調処理に関する『北見』サマー・ワークショップ (SWoPP2019)
    • Related Report
      2019 Annual Research Report
  • [Presentation] 相対的高メモリバンド幅環境における密行列固有値ソルバの実装方式について2019

    • Author(s)
      片桐孝洋
    • Organizer
      日本応用数理学会2019年度年会
    • Related Report
      2019 Annual Research Report
  • [Presentation] Initial particle distribution based on the centroidal Voronoi tessellation for two-dimensional particle method2019

    • Author(s)
      Hayate Hasegawa, Masao Ogino, Takahiro Katagiri
    • Organizer
      The 7th Asia-Pacific Congress on Computational MechanicsProgram (APCOM2019)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] ppOpen-ATによる自動チューニングの実行2019

    • Author(s)
      関直人
    • Organizer
      第20回AT研究会オープンアカデミックセッション(ATOS20)
    • Related Report
      2019 Annual Research Report
    • Invited
  • [Presentation] Enhancement of Performance Parameter Search Method for Multiple Parameter Estimation2019

    • Author(s)
      Teruo Tanaka, Fan Guuing, Akihiro Fujii, Takahiro Katagiri
    • Organizer
      Conference on Advanced Topics and Auto Tuning in High-Performance Scientific Computing (ATAT in HPSC 2019)
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research
  • [Presentation] ソフトウェア自動チューニングにおける複数性能パラメータを同時推定する手法の効率化2019

    • Author(s)
      多部田敏樹,田中輝雄, 藤井昭宏,関 直人,范 谷瑛
    • Organizer
      第81回情報処理学会全国大会
    • Related Report
      2018 Research-status Report
  • [Presentation] 共役勾配法におけるダブルバッファリング利用したRDMA通信の性 能評価2019

    • Author(s)
      出蔵英真,藤井昭宏,田中輝雄
    • Organizer
      第81回情報処理学会全国大会
    • Related Report
      2018 Research-status Report
  • [Presentation] Toward Auto-tuning of Preconditioners for Sparse Iterative Solvers by Deep Learning2019

    • Author(s)
      Takahiro Katagiri
    • Organizer
      2019 Conference on Advanced Topics and Auto Tuning in High-Performance Scientific Computing (ATAT2019)
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research
  • [Presentation] Auto-tuning of Preconditioners with Deep Learning2019

    • Author(s)
      Takahiro Katagiri, Kenya Yamada
    • Organizer
      SIAM Conference on Computational Science and Engineering (CSE19)
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research
  • [Presentation] Optimization of hierarchical matrix computation on GPU --- accelerating many small matrix calculation2018

    • Author(s)
      Satoshi Ohshima, Ichitaro Yamazaki, Akihiro Ida, Rio Yokota
    • Organizer
      Sapporo Summer HPC Seminar 2018
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research
  • [Presentation] マルチコア・メニーコア計算機環境における Chebyshev基底通信削減CG法の性能評価2018

    • Author(s)
      大島聡史, 藤井昭宏, 田中輝雄, 深谷猛, 須田礼仁
    • Organizer
      日本応用数理学会2018年 年会
    • Related Report
      2018 Research-status Report
  • [Remarks] Xabclib (eXteneded ABCLib) Project

    • URL

      http://www.abc-lib.org/Xabclib/

    • Related Report
      2019 Annual Research Report
  • [Funded Workshop] The Fourteenth International Workshop on Automatic Performance Tuning (iWAPT2019)2019

    • Related Report
      2019 Annual Research Report

URL: 

Published: 2018-07-25   Modified: 2021-02-19  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi