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

Breakthrough in fundamental technology for ultralow-power neuromorphic hardware

Research Project

Project/Area Number 21H04887
Research Category

Grant-in-Aid for Scientific Research (A)

Allocation TypeSingle-year Grants
Section一般
Review Section Medium-sized Section 61:Human informatics and related fields
Research InstitutionThe University of Tokyo

Principal Investigator

KOHNO Takashi  東京大学, 生産技術研究所, 教授 (90447350)

Co-Investigator(Kenkyū-buntansha) 小林 正治  東京大学, 大学院工学系研究科(工学部), 准教授 (40740147)
Project Period (FY) 2021-04-05 – 2024-03-31
Project Status Completed (Fiscal Year 2023)
Budget Amount *help
¥42,250,000 (Direct Cost: ¥32,500,000、Indirect Cost: ¥9,750,000)
Fiscal Year 2023: ¥13,780,000 (Direct Cost: ¥10,600,000、Indirect Cost: ¥3,180,000)
Fiscal Year 2022: ¥13,650,000 (Direct Cost: ¥10,500,000、Indirect Cost: ¥3,150,000)
Fiscal Year 2021: ¥14,820,000 (Direct Cost: ¥11,400,000、Indirect Cost: ¥3,420,000)
Keywordsニューロモルフィックハードウェア / 神経模倣システム / 低電力アナログ集積回路 / 教師なし学習 / 超低電力アナログ回路 / 神経模倣回路 / 神経模倣情報処理 / 超低消費電力アナログ回路 / 不揮発性メモリデバイス / シリコン神経ネットワーク / ニューロモルフィック回路
Outline of Research at the Start

脳からより多くを学ぶことで脳により近い情報処理の実現を目指すニューロモルフィックハードウェアにおける重要課題(超低電力アナログ集積回路実装技術と、より脳に近い情報処理モデル)を融合的に研究する。これによって、現行の人工知能(AI)の限界を超え、ヒトの脳のように自発的で複雑な処理を超低電力で行う次世代AIのための、CMOS/FeFET混在アナログシリコン神経ネットワーク基盤技術を進展させる。

Outline of Final Research Achievements

Fundamental technologies for enhancing power efficiency of neuromorphic hardware were developed by being aware of detailed mechanisms of information processing in the nervous system. Establishing the fundamental technology for neuromorphic hardware with power efficiency comparable to the nervous system is included in the scope of this work. By reproducing the dynamical structures in the neuronal activities, we developed an ultra-low power (~200 pW) neuron circuit. In addition, unsupervised learning models that exploit noises and can be implemented efficiently by mixed-signal circuits were developed.

Academic Significance and Societal Importance of the Research Achievements

ニューロモルフィックハードウェアは、神経スパイクと呼ばれるパルス状の電気活動が脳神経ネットワークの情報処理の重要な要素であることから着想された、パルスを用いて情報をコーディングする超並列ハードウェアであり、人工知能(AI)と同等の情報処理を低電力で実行できるハードウェア基盤として注目されている。本研究では、より脳神経系から学ぶことで、脳神経ネットワークに近いエネルギー効率で動作するニューロモルフィックハードウェアの基礎技術を開発した。本技術は、現行のニューロモルフィックハードウェアよりエネルギー効率が3桁程度高い新しいニューロモルフィックハードウェアの基盤技術となりうる。

Report

(5 results)
  • 2023 Annual Research Report   Final Research Report ( PDF )
  • 2022 Annual Research Report
  • 2021 Comments on the Screening Results   Annual Research Report
  • Research Products

    (9 results)

All 2024 2023 2022 2021

All Journal Article (3 results) (of which Peer Reviewed: 3 results,  Open Access: 3 results) Presentation (6 results) (of which Int'l Joint Research: 6 results)

  • [Journal Article] Adaptive STDP-based on-chip spike pattern detection2023

    • Author(s)
      Gautam Ashish、Kohno Takashi
    • Journal Title

      Frontiers in Neuroscience

      Volume: 17 Pages: 1-15

    • DOI

      10.3389/fnins.2023.1203956

    • Related Report
      2023 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] A Conductance-Based Silicon Synapse Circuit2022

    • Author(s)
      Gautam Ashish、Kohno Takashi
    • Journal Title

      Biomimetics

      Volume: 7 Issue: 4 Pages: 246-246

    • DOI

      10.3390/biomimetics7040246

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] An Adaptive STDP Learning Rule for Neuromorphic Systems2021

    • Author(s)
      Gautam Ashish、Kohno Takashi
    • Journal Title

      Frontiers in Neuroscience

      Volume: 15 Pages: 1-12

    • DOI

      10.3389/fnins.2021.741116

    • Related Report
      2021 Annual Research Report
    • Peer Reviewed / Open Access
  • [Presentation] Neuromorphic spatio-temporal spike pattern detection model with biological plausibility2024

    • Author(s)
      Shunta Furuichi and Takashi Kohno
    • Organizer
      The 12th RIEC International Symposium on Brain Functions and Brain Computer
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Spike pattern learning through STDP mechanisms with competitive dynamics and axonal delays2024

    • Author(s)
      Zhaoyu Hao
    • Organizer
      The 12th RIEC International Symposium on Brain Functions and Brain Computer
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Parameter fitting approach for the piecewise quadratic neuron model using improved particle swarm optimization framework2024

    • Author(s)
      Zihan Yang and Takashi Kohno
    • Organizer
      The 12th RIEC International Symposium on Brain Functions and Brain Computer
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] On-chip spike pattern classification for neuromorphic systems2024

    • Author(s)
      Ashish Gautam and Takashi Kohno
    • Organizer
      American Physical Society March Meeting 2024
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Toward on-chip STDP learning on mixed-signal neuromorphic chips2023

    • Author(s)
      Ashish Gautam, Takashi Kohno, and Prasanna Date
    • Organizer
      ICRC 2023: THE 8TH IEEE INTERNATIONAL CONFERENCE ON REBOOTING COMPUTING (ICRC) 2023
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Adaptive STDP Learning with Lateral Inhibition for Neuromorphic Systems2023

    • Author(s)
      Ashish Gautam, Takashi Kohno
    • Organizer
      The 2023 International Conference on Artificial Life and Robotics
    • Related Report
      2022 Annual Research 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