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

A Study of Evidence-Based Performance Tuning

Research Project

Project/Area Number 26540031
Research Category

Grant-in-Aid for Challenging Exploratory Research

Allocation TypeMulti-year Fund
Research Field Software
Research InstitutionChiba Institute of Technology (2017)
Institute of Physical and Chemical Research (2014-2016)

Principal Investigator

HASHIMOTO Masatomo  千葉工業大学, 人工知能・ソフトウェア技術研究センター, 上席研究員 (60357770)

Project Period (FY) 2014-04-01 – 2018-03-31
Project Status Completed (Fiscal Year 2017)
Budget Amount *help
¥3,770,000 (Direct Cost: ¥2,900,000、Indirect Cost: ¥870,000)
Fiscal Year 2016: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2015: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2014: ¥1,950,000 (Direct Cost: ¥1,500,000、Indirect Cost: ¥450,000)
Keywords高性能計算 / 性能チューニング / 最適化パターン推定 / 計算カーネル同定 / 目標実行効率推定 / プログラム構造理解支援 / プログラム解析 / 機械学習 / 計算カーネル分類 / 実行効率予測 / 大規模調査 / コード理解支援 / アプリケーション性能チューニング / 大規模科学技術計算 / 計算カーネル特定 / 最適化パターン予測 / ソースコード変更パターン同定
Outline of Final Research Achievements

Performance tuning is still a demanding manual task. To improve an application's efficiency, we have to identify its computational kernels, each of which is typically composed of one or more loops. Then various empirical attempts such as loop transformations are made. Thus, it is crucial to learn from the experience of performance tuning experts. As a proof-of-concept, we extracted various facts from performance tuning histories of a few real-world scientific applications, and then constructed a database of that facts. Based on the database, we constructed a few experimental predictive models for promising loop transformation patterns. We also explored a thousand computation-intensive applications to reveal the distribution of kernel classes, each of which is related to expected efficiency and specific tuning patterns. In addition, we constructed a binary classifier for identifying loop kernels and a multi-class classifier for predicting kernel classes.

Report

(5 results)
  • 2017 Annual Research Report   Final Research Report ( PDF )
  • 2016 Research-status Report
  • 2015 Research-status Report
  • 2014 Research-status Report
  • Research Products

    (10 results)

All 2018 2017 2015 2014 Other

All Journal Article (1 results) (of which Peer Reviewed: 1 results,  Acknowledgement Compliant: 1 results) Presentation (4 results) (of which Int'l Joint Research: 2 results,  Invited: 1 results) Remarks (5 results)

  • [Journal Article] An Empirical Study of Computation-Intensive Loops for Identifying and Classifying Loop Kernels2017

    • Author(s)
      Masatomo Hashimoto, Masaaki Terai, Toshiyuki Maeda, and Kazuo Minami
    • Journal Title

      Proceedings of the 8th ACM/SPEC on International Conference on Performance Engineering

      Volume: - Pages: 361-372

    • DOI

      10.1145/3030207.3030217

    • Related Report
      2016 Research-status Report
    • Peer Reviewed / Acknowledgement Compliant
  • [Presentation] CCA/EBT: Code Comprehension Assistance Tool for Evidence-Based Performance Tuning2018

    • Author(s)
      Masatomo Hashimoto, Toshiyuki Maeda
    • Organizer
      International Conference on High Performance Computing in Asia-Pacific Region
    • Related Report
      2017 Annual Research Report
    • Int'l Joint Research
  • [Presentation] An Empirical Study of Computation-Intensive Loops for Identifying and Classifying Loop Kernels2017

    • Author(s)
      Masatomo Hashimoto
    • Organizer
      8th ACM/SPEC on International Conference on Performance Engineering
    • Related Report
      2017 Annual Research Report 2016 Research-status Report
    • Int'l Joint Research
  • [Presentation] Extracting Facts from Performance Tuning History of Scientific Applications for Predicting Effective Optimization Patterns2015

    • Author(s)
      Masatomo Hashimoto
    • Organizer
      The 12th Working Conference on Mining Software Repositories
    • Place of Presentation
      Florence, Italy
    • Year and Date
      2015-05-16 – 2015-05-17
    • Related Report
      2014 Research-status Report
  • [Presentation] 根拠に基づく性能チューニングの支援に向けて2014

    • Author(s)
      橋本政朋
    • Organizer
      第6回 自動チューニング技術の現状と応用に関するシンポジウム (ATTA2014)
    • Place of Presentation
      東京大学 山上会館
    • Year and Date
      2014-12-25
    • Related Report
      2014 Research-status Report
    • Invited
  • [Remarks] 事例に基づくプログラム性能チューニング支援

    • URL

      https://stair.center/archives/project/ebt

    • Related Report
      2017 Annual Research Report
  • [Remarks] (1) CCA/EBT Experiments

    • URL

      https://github.com/ebt-hpc/icpe2017

    • Related Report
      2016 Research-status Report
  • [Remarks] (2) CCA/EBT

    • URL

      https://github.com/ebt-hpc/cca

    • Related Report
      2016 Research-status Report
  • [Remarks] (4) Docker Image of CCA/EBT

    • URL

      https://hub.docker.com/r/ebtxhpc/cca/

    • Related Report
      2016 Research-status Report
  • [Remarks] (3) docker-cca

    • URL

      https://github.com/ebt-hpc/docker-cca

    • Related Report
      2016 Research-status Report

URL: 

Published: 2014-04-04   Modified: 2019-03-29  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi