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

Write latency reduction on PCM for approximate computing

Research Project

Project/Area Number 20K11728
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 60040:Computer system-related
Research InstitutionChiba University

Principal Investigator

Kazuteru NAMBA  千葉大学, 大学院工学研究院, 准教授 (60359594)

Co-Investigator(Kenkyū-buntansha) イン ユウ  群馬大学, 大学院理工学府, 教授 (10520124)
Project Period (FY) 2020-04-01 – 2023-03-31
Project Status Completed (Fiscal Year 2022)
Budget Amount *help
¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Fiscal Year 2022: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2021: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2020: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Keywordsメモリシステム / 誤差許容計算 / ニューラルネット / 低電力化 / 高信頼化 / 相変化メモリ / 計算機システム
Outline of Research at the Start

新しいメモリシステムである PCM (相変化メモリとも呼ばれる) について,その平均書き込み時間削減を目的とする.PCM の書き込み時間削減は申請者らが切り開きつつある計算機システムにおける新しい分野の1種であり学術的興味も高い.本申請課題では特に誤差許容計算実行時の書き込み時間削減について考える.誤差許容計算 は人工知能への使用など要求が高まっている技術であり,その工業的産業的重要性は高い.

Outline of Final Research Achievements

This work has not achieved a write time reduction, which was the main objective. However, we have presented several related techniques, such as power consumption reduction on memory systems for an approximate computing system.
For example, we have presented a power consumption reduction for a neural network system, a typical example of an approximate computing system. The proposed system uses two different power supply voltages. The proposed method achieves a power consumption reduction of 35%, avoiding a significant reduction in the recognition rate.

Academic Significance and Societal Importance of the Research Achievements

本研究においては主目的であった書き込み時間削減については結果を出せていない。しかし,副産物と言える消費電力削減手法などはいずれも実用性の高いものであり,本研究成果の工業的産業的重要性は十分に高いものであったと言える。また,目的であったメモリシステム書き込み時間についてもいくつかの知見が得られており,研究当初に考えていた問いにもいくらかは答えられた,学術的にも意義がある研究であったと考えている。

Report

(4 results)
  • 2022 Annual Research Report   Final Research Report ( PDF )
  • 2021 Research-status Report
  • 2020 Research-status Report
  • Research Products

    (12 results)

All 2023 2022 2021 2020 Other

All Int'l Joint Research (1 results) Journal Article (1 results) (of which Peer Reviewed: 1 results,  Open Access: 1 results) Presentation (10 results) (of which Int'l Joint Research: 3 results)

  • [Int'l Joint Research] Pusan National University(韓国)

    • Related Report
      2022 Annual Research Report
  • [Journal Article] Low Power Neural Network by Reducing SRAM Operating Voltage2022

    • Author(s)
      Kozu Keisuke、Tanabe Yuya、Kitakami Masato、Namba Kazuteru
    • Journal Title

      IEEE Access

      Volume: 10 Pages: 116982-116986

    • DOI

      10.1109/access.2022.3219208

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access
  • [Presentation] 外れ値を用いたDNNの縮退故障に対するエラー耐性の向上2023

    • Author(s)
      石井 智大, 難波 一輝
    • Organizer
      IEICE FIIS
    • Related Report
      2022 Annual Research Report
  • [Presentation] Stuck-at Fault Tolerance in DNN Using Outliers and Sampling2023

    • Author(s)
      Tomohiro Ishii, Donghyun Kwon and Kazuteru Namba
    • Organizer
      Japan-Korea Joint Workshop on Complex Communication Sciences
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Stuck-at fault tolerance in DNN using statistical data2022

    • Author(s)
      Tomohiro Ishii and Kazuteru Namba
    • Organizer
      IEEE Pacific Rim International Symposium on Dependable Computing
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] 外れ値と標本化を用いたDNNの縮退故障に対するエラー耐性の向上2022

    • Author(s)
      石井 智大, 難波 一輝
    • Organizer
      IEICE DC
    • Related Report
      2022 Annual Research Report
  • [Presentation] 動作電圧引き下げによる低消費電力ニューラルネットワークのための6T-8TハイブリッドSRAM2022

    • Author(s)
      余 若曦, 難波 一輝
    • Organizer
      IEICE DC
    • Related Report
      2022 Annual Research Report
  • [Presentation] Low power quantized neural network by reducing the operating voltage of SRAM2022

    • Author(s)
      Ji Wu, Kazuteru Namba
    • Organizer
      IEICE DC
    • Related Report
      2022 Annual Research Report
  • [Presentation] Relaxing device requirements for non-linearity in Deep Neural Networks accelerators with Phase Change Memory2021

    • Author(s)
      Keisuke Kozu and Kazuteru Namba
    • Organizer
      IEEE Int'l Conf. Consum. Electron. Taiwan
    • Related Report
      2021 Research-status Report
    • Int'l Joint Research
  • [Presentation] SRAMの動作電圧引き下げによるニューラルネットワークの低電力化2021

    • Author(s)
      高津 啓佑, 難波 一輝
    • Organizer
      信学技報, DC
    • Related Report
      2021 Research-status Report
  • [Presentation] マルチレベルセル相変化メモリを用いた連想メモリ2021

    • Author(s)
      高橋 知宏, 難波一輝
    • Organizer
      信学技報, DC
    • Related Report
      2021 Research-status Report
  • [Presentation] 相変化メモリを用いた赤黒木構造の書き込み時間削減2020

    • Author(s)
      楊 昊天, 難波 一輝
    • Organizer
      信学技報, FIIS
    • Related Report
      2020 Research-status Report

URL: 

Published: 2020-04-28   Modified: 2024-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi