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

FPGA-driven Broadband Optical Communication Interconnect Network

Research Project

Project/Area Number 21K11859
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 60060:Information network-related
Research InstitutionKeio University (2022-2023)
National Institute of Informatics (2021)

Principal Investigator

HU Yao  慶應義塾大学, デジタルメディア・コンテンツ統合研究センター(日吉), 特任助教 (50791232)

Project Period (FY) 2021-04-01 – 2024-03-31
Project Status Completed (Fiscal Year 2023)
Budget Amount *help
¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Fiscal Year 2023: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2022: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2021: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Keywords高性能計算 / 計算機ネットワーク / データ圧縮 / タスクスケジューリング / 大規模計算機システム / 光通信ネットワーク / FPGA結合網 / 広帯域光通信 / データセンターネットワーク
Outline of Research at the Start

あらゆるモノがデータを生み出す「データ中心型社会」の時代が到来している。ムーアの法則に従った汎用プロセッサの「計算」の性能向上が望めない中で、データセンター内の「通信」が性能向上の鍵となる。本研究では、異なるハードウェアを直接結合し通信やスケジューリングのオーバヘッドを削減したうえで、広帯域光通信技術を用いた異種ハードウェア混在環境のデータセンターにおける低遅延かつ高効率な処理を探求する。そこで、様々な並列アプリケーションを1つのデータセンターで低遅延かつ効率的にサポートすることが期待できる。

Outline of Final Research Achievements

In this study, we proposed a dynamic configuration method for FPGA-driven optical communication topologies tailored to the input communication patterns, and investigated a design approach for datacenter systems that can also accommodate post-Moore architectures. This enables the reduction of overall power consumption and communication volume in datacenters, as well as the number of hops (communication delay) between computing nodes as the computer systems scale up. In addition, we developed and utilized dynamic configuration methods for optical communication topologies, and proposed a flexible job scheduling method for parallel applications. We also examined application-level techniques for approximate communication to enable high portability on high-performance interconnection networks. This is expected to minimize waiting time and total execution time for parallel applications while maximizing overall system scalability.

Academic Significance and Societal Importance of the Research Achievements

本研究で開発したトポロジ動的構成法、データ圧縮法とスパコンスケジューラのプログラムをオープンソースソフトウェアとして公開した。研究過程で得られた知見については、産業界・学術界の技術者・研究者らと幅広い議論を交えながら、研究会・国際会議・論文誌などで発表し、将来のスパコンネットワーク設計や異種資源環境データセンター構築に向けた参考とする。
本研究により、異種資源環境システムにおける大規模計算機ネットワークがそのポテンシャルを十分に発揮することで、ビッグデータ時代のニューラルネットワークにおける多様なワークロード処理の速度や次世代アプリケーション実行性能をより一層向上させることが期待できる。

Report

(4 results)
  • 2023 Annual Research Report   Final Research Report ( PDF )
  • 2022 Research-status Report
  • 2021 Research-status Report
  • Research Products

    (7 results)

All 2023 2022 2021

All Journal Article (2 results) (of which Peer Reviewed: 2 results) Presentation (5 results) (of which Int'l Joint Research: 3 results)

  • [Journal Article] Accelerating Parallel Applications Based on Graph Reordering for Random Network Topologies2023

    • Author(s)
      Hu Yao
    • Journal Title

      IEEE Access

      Volume: 11 Pages: 40373-40383

    • DOI

      10.1109/access.2023.3269793

    • Related Report
      2023 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Exploring Approximate Communication Using Lossy Bitwise Compression on Interconnection Networks2023

    • Author(s)
      Hu Yao
    • Journal Title

      IEEE Access

      Volume: 11 Pages: 59238-59249

    • DOI

      10.1109/access.2023.3281834

    • Related Report
      2023 Annual Research Report
    • Peer Reviewed
  • [Presentation] Optimizing Link Prediction by Community Detection on Dynamic Networks2023

    • Author(s)
      Hu Yao、Huang Qian
    • Organizer
      2023 IEEE 11th International Conference on Information, Communication and Networks (ICICN)
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Parallel Computation of Random Walks on Distributed Graphs2023

    • Author(s)
      Hu Yao
    • Organizer
      26th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD2023)
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] 並列計算機における自動チューニングを用いた通信最適化2022

    • Author(s)
      胡曜
    • Organizer
      ETNET2022
    • Related Report
      2022 Research-status Report
  • [Presentation] 非可逆圧縮を用いたMPI通信の性能評価2022

    • Author(s)
      胡曜
    • Organizer
      HotSPA2022
    • Related Report
      2022 Research-status Report
  • [Presentation] MiniCAR: Minimal Congestion-Aware Routing Method in Fine-Grained Circuit-Switched Networks for Parallel Computing Systems2021

    • Author(s)
      Yao Hu
    • Organizer
      26th IEEE Symposium on Computers and Communications (ISCC 2021)
    • Related Report
      2021 Research-status Report
    • Int'l Joint Research

URL: 

Published: 2021-04-28   Modified: 2025-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi