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

Real-Time Data Kernel for Data Intensive Science

Research Project

Project/Area Number 19H04117
Research Category

Grant-in-Aid for Scientific Research (B)

Allocation TypeSingle-year Grants
Section一般
Review Section Basic Section 60080:Database-related
Research InstitutionKeio University

Principal Investigator

KAWASHIMA Hideyuki  慶應義塾大学, 環境情報学部(藤沢), 准教授 (90407148)

Co-Investigator(Kenkyū-buntansha) 松谷 宏紀  慶應義塾大学, 理工学部(矢上), 准教授 (70611135)
藤原 靖宏  日本電信電話株式会社NTTコミュニケーション科学基礎研究所, 上田特別研究室, 主任研究員 (70837971)
Project Period (FY) 2019-04-01 – 2022-03-31
Project Status Completed (Fiscal Year 2021)
Budget Amount *help
¥17,680,000 (Direct Cost: ¥13,600,000、Indirect Cost: ¥4,080,000)
Fiscal Year 2021: ¥2,860,000 (Direct Cost: ¥2,200,000、Indirect Cost: ¥660,000)
Fiscal Year 2020: ¥5,460,000 (Direct Cost: ¥4,200,000、Indirect Cost: ¥1,260,000)
Fiscal Year 2019: ¥9,360,000 (Direct Cost: ¥7,200,000、Indirect Cost: ¥2,160,000)
Keywordsトランザクション / 機械学習 / アクセラレータ / データベース / ロギング / 不揮発メモリ / データシステム / 障害回復 / アナリティクス
Outline of Research at the Start

すばる望遠鏡の天体カタログをモデルケースとして、本研究ではデータ集約型科学に資する新しいデータカーネルの実現法を探求する。本研究の第一の独自性は、性能評価指標のスループットからリアルタイム性への転換である。近年はスループットのみが評価されるが、天体カタログでは望遠鏡の即時制御のためにリアルタイム性が重要である。本研究の第二の独自性は、システム構造の簡素化から稠密化への転換である。近年のデータカーネルは集積化傾向があるが、天体カタログが求める全3 要素を集積する程の高稠密化法は、未開拓段階にある。本研究により、我が国の天文研究の世界トップ維持を支援しつつ、他データ集約型科学への展開を狙う。

Outline of Final Research Achievements

In order to create a real-time data kernel for data-intensive science, we conducted research on high-performance transaction processing, fast machine learning algorithms, and fast accelerators for AI. For transaction processing, we developed CCBench, a platform that enables comprehensive comparison and evaluation of modern methods, explored superior protocols, and devised optimization methods. We also applied these methods to robotics. For machine learning, we accelerated key algorithms such as b-matching and anchor graphs. For accelerators, we have succeeded in improving the performance of DQN and other algorithms using FPGAs.

Academic Significance and Societal Importance of the Research Achievements

本研究ではリアルタイムデータカーネルの創出を目的とし,トランザクション処理というコア技術を研磨した結果,ロボット用ミドルウェアROSをリアルタイム化かつ正確化する技術を開発し,また,それをソフトウェアとして創出するに至った.このような成果は我々の知る限り存在しない.機械学習アルゴリズムならびにアクセラレータについては,SOTAに優る成果を生み出せた.

Report

(4 results)
  • 2021 Annual Research Report   Final Research Report ( PDF )
  • 2020 Annual Research Report
  • 2019 Annual Research Report
  • Research Products

    (9 results)

All 2022 2021 2020

All Journal Article (2 results) (of which Peer Reviewed: 2 results,  Open Access: 2 results) Presentation (7 results) (of which Int'l Joint Research: 7 results)

  • [Journal Article] Scalable Distributed Metadata Server Based on Nonblocking Transactions2020

    • Author(s)
      Kohei Hiraga, Osamu Tatebe, Hideyuki Kawashima
    • Journal Title

      The Journal of Universal Computer Science

      Volume: 26 Pages: 89-106

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] An Analysis of Concurrency Control Protocols for In-Memory Database with CCBench2020

    • Author(s)
      Takayuki Tanabe, Takashi Hoshino, Hideyuki Kawashima, Osamu Tatebe
    • Journal Title

      Proceedings of the VLDB Endowment

      Volume: 13 Issue: 13 Pages: 3531-3534

    • DOI

      10.14778/3424573.3424575

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed / Open Access
  • [Presentation] dsODENet: Neural ODE and Depthwise Separable Convolution for Domain Adaptation on FPGAs2022

    • Author(s)
      Hiroki Kawakami, Hirohisa Watanabe, Keisuke Sugiura, Hiroki Matsutani
    • Organizer
      30th Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP'22)
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Accelerating Concurrency Control with Active Thread Adjustment2022

    • Author(s)
      Kosei Masumura, Takashi Hoshino, Hideyuki Kawashima
    • Organizer
      IEEE BigComp
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Stable Low Latency Logging for Epoch-based In-memory Database.2022

    • Author(s)
      Masahiro Tanaka, Hideyuki Kawashima.
    • Organizer
      IEEE BigComp
    • Related Report
      2020 Annual Research Report
    • Int'l Joint Research
  • [Presentation] An FPGA-Based On-Device Reinforcement Learning Approach using Online Sequential Learning2021

    • Author(s)
      Hirohisa Watanabe, Mineto Tsukada, Hiroki Matsutani
    • Organizer
      35th IEEE International Parallel and Distributed Processing Symposium (IPDPS'21) Workshops
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Fast and Accurate Anchor Graph-based Label Prediction2021

    • Author(s)
      Yasuhiro Fujiwara, Yasutoshi Ida, Atsutoshi Kumagai, Sekitoshi Kanai, Naonori Ueda
    • Organizer
      The Conference on Information and Knowledge Management
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Efficient Algorithm for the b-Matching Graph2020

    • Author(s)
      Yasuhiro Fujiwara, Atsutoshi Kumagai, Sekitoshi Kanai, Yasutoshi Ida, Naonori Ueda
    • Organizer
      The 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] An Edge Attribute-wise Partitioning and Distributed Processing of R-GCN using GPUs2020

    • Author(s)
      Tokio Kibata, Mineto Tsukada, Hiroki Matsutani
    • Organizer
      26th International European Conference on Parallel and Distributed Computing (Euro-Par'20) Workshops (HeteroPar'20)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research

URL: 

Published: 2019-04-18   Modified: 2023-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi