東北大学研究者紹介

タカハシ　ケイチ

髙橋　慧智

Keichi Takahashi

所属

サイバーサイエンスセンター　研究開発部　スーパーコンピューティング研究部

職名

助教

学位

博士（情報科学）（大阪大学）
修士（情報科学）（大阪大学）

researchmap

https://researchmap.jp/keichi

J-GLOBAL ID

201901008336691253

e-Rad 研究者番号

40846408

プロフィール

高性能計算およびネットワーキング技術に関する研究に従事．次世代の大規模な高性能計算機におけるストレージI/Oや通信の高速化に興味を持つ．米Oak Ridge国立研究所や米Salk研究所などの海外研究機関と国際連携研究を積極的に推進している．学生時代より複数の企業においてソフトウェア開発に携わってきた経験を有し，現在も多数のオープンソースソフトウェアの開発に貢献している．

経歴 6

2021年12月～継続中

東北大学　サイバーサイエンスセンター　助教
2019年4月～ 2021年11月

奈良先端科学技術大学院大学　助教
2022年1月～継続中

奈良先端科学技術大学院大学　客員助教
2019年6月～ 2022年3月

大阪大学　サイバーメディアセンター　招聘教員
2020年4月～ 2021年10月

奈良工業高等専門学校　非常勤講師
2018年7月～ 2018年12月

米国オークリッジ国立研究所　科学データグループ　客員研究員

︎全件表示 ︎最初の5件までを表示

学歴 2

大阪大学　大学院情報科学研究科

2014年4月～ 2019年3月
大阪大学　工学部

2010年4月～ 2014年3月

委員歴 14

HPCI 連携サービス運営・作業部会　部会員

2023年4月～継続中
電子情報通信学会　常任査読委員

2022年7月～ 2023年7月
電子情報通信学会　英文論文誌D編集委員

2019年6月～ 2023年5月
情報処理学会関西支部　支部幹事

2020年5月～ 2022年5月
情報処理学会関西支部　2021年度支部大会実行委員長

2021年4月～ 2022年3月
APNOMS Technical Program Committee

2019年～ 2021年
SC24　Poster Committee

2024年～
HPCAsia 2024 Poster Chair

2024年～
PDCAT 2023　Program Committee

2023年～
IEEE CLUSTER 2023　Program Committee

2023年～
HiPC 2023 Program Committee

2023年～
SC22　Program Committee

2022年～
PDCAT 2022　Publicity Chair

2022年～
xSIG 2021 プログラム委員

2021年～

︎全件表示 ︎最初の5件までを表示

所属学協会 3

IEEE Computer Society
情報処理学会
IEEE

研究分野 2

情報通信 / 情報ネットワーク /
情報通信 / 高性能計算 /

論文 60

Parallelizable Loop Detection using Pre-trained Transformer Models for Code Understanding 査読有り

Soratouch Pornmaneerattanatri, Keichi Takahashi, Yutaro Kashiwa, Kohei Ichikawa, Hajimu Iida

Parallel and Distributed Computing, Applications and Technologies　32-42　2023年11月29日
出版者・発行元：Springer Nature Singapore
DOI： 10.1007/978-981-99-8211-0_4 　

ISSN：1876-1100

eISSN：1876-1119
Multi-Objective Optimization of Controller Placement in Distributed ONOS Networks 査読有り

Xingyuan Kang, Keichi Takahashi, Chawanat Nakasan, Kohei Ichikawa, Hajimu Iida

The Eleventh International Symposium on Computing and Networking (CANDAR 2023)　2023年11月
ベクトル型スーパーコンピュータ「AOBA-S」の性能評価

高橋慧智, 藤本壮也, 長瀬悟, 磯部洋子, 下村陽一, 江川隆輔, 滝沢寛之

研究報告ハイパフォーマンスコンピューティング（HPC）　2023-HPC-191　(1)　1-9　2023年9月
Comparison of Parallel STL with C/C++ GPU Programming Models

Joanna Imada, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

研究報告ハイパフォーマンスコンピューティング（HPC）　2023-HPC-190　(2)　1-7　2023年7月
The Convergence of Container and Traditional Virtualization: Strengths and Limitations 査読有り

Guoqing Li, Keichi Takahashi, Kohei Ichikawa, Hajimu Iida, Chawanat Nakasan, Pattara Leelaprute, Pree Thiengburanathum, Passakorn Phannachitta

SN Computer Science　4　(4)　2023年5月11日
出版者・発行元：Springer Science and Business Media LLC
DOI： 10.1007/s42979-023-01827-9 　

eISSN：2661-8907
Balancing exploitation and exploration in parallel Bayesian optimization under computing resource constraint

Moto Satake, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

2023 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)　2023年5月
出版者・発行元：IEEE
DOI： 10.1109/ipdpsw59300.2023.00122 　
Equivalence Checking of Code Transformation by Numerical and Symbolic Approaches 査読有り

Shunpei Sugawara, Keichi Takahashi, Yoichi Shimomura, Ryusuke Egawa, Hiroyuki Takizawa

Parallel and Distributed Computing, Applications and Technologies　373-386　2023年4月8日
出版者・発行元：Springer Nature Switzerland
DOI： 10.1007/978-3-031-29927-8_29 　

ISSN：0302-9743

eISSN：1611-3349
Towards Priority-Flexible Task Mapping for Heterogeneous Multi-core NUMA Systems 査読有り

Yifan Jin, Mulya Agung, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

Parallel and Distributed Computing, Applications and Technologies　3-15　2023年4月8日
出版者・発行元：Springer Nature Switzerland
DOI： 10.1007/978-3-031-29927-8_1 　

ISSN：0302-9743

eISSN：1611-3349
A Task-Parallel Runtime for Heterogeneous Multi-node Vector Systems 査読有り

Kazuki Ide, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

Parallel and Distributed Computing, Applications and Technologies　331-343　2023年4月8日
出版者・発行元：Springer Nature Switzerland
DOI： 10.1007/978-3-031-29927-8_26 　

ISSN：0302-9743

eISSN：1611-3349
探索と活用の調整による並列ベイズ最適化の効率化

佐竹望都, 高橋慧智, 下村陽一, 滝沢寛之

研究報告ハイパフォーマンスコンピューティング（HPC）　2023-HPC-188　(33)　1-8　2023年3月
エッジコンピューティングにおける拠点間の通信遅延を考慮したリクエスト分散制御の評価

村上恭哉, 高橋慧智, 市川昊平, 飯田元

信学技報　2023年2月
Prototype of a Batched Quantum Circuit Simulator for the Vector Engine. 査読有り

Keichi Takahashi, Toshio Mori, Hiroyuki Takizawa

SC Workshops　1499-1505　2023年

DOI： 10.1145/3624062.3624226 　
Efficient Pause Location Prediction Using Quantum Annealing Simulations and Machine Learning. 査読有り

Michael R. Zielewski, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

IEEE Access　11　104285-104294　2023年

DOI： 10.1109/ACCESS.2023.3317698 　
Scalable Empirical Dynamic Modeling With Parallel Computing and Approximate k-NN Search. 査読有り

Keichi Takahashi, Kohei Ichikawa, Joseph Park, Gerald M. Pao

IEEE Access　11　68171-68183　2023年

DOI： 10.1109/ACCESS.2023.3289836 　
Toward Predictive Modeling of Solar Power Generation for Multiple Power Plants 査読有り

Thonglek, K., Ichikawa, K., Takahashi, K., Nakasan, C., Yuasa, K., Babasaki, T., Iida, H.

IEICE Transactions on Communications　E106.B　(7)　2023年

DOI： 10.1587/transcom.2022EBT0003 　

ISSN：1745-1345 0916-8516
Performance Evaluation of a Next-Generation SX-Aurora TSUBASA Vector Supercomputer 査読有り

Keichi Takahashi, Soya Fujimoto, Satoru Nagase, Yoko Isobe, Yoichi Shimomura, Ryusuke Egawa, Hiroyuki Takizawa

ISC 2023　359-378　2023年

DOI： 10.1007/978-3-031-32041-5_19 　
A Real-time Flood Inundation Prediction on SX-Aurora TSUBASA 査読有り

Yoichi Shimomura, Akihiro Musa, Yoshihiko Sato, Atsuhiko Konja, Guoqing Cui, Rei Aoyagi, Keichi Takahashi, Hiroyuki Takizawa

29th International Conference on High Performance Computing, Data, and Analytics (HiPC)　2022年12月
蓄電システムのロバスト制御によるマルチユース実証試験～需要・発電予測精度評価～

植嶋美喜, 湯淺一史, 竹内義晴, 馬場﨑忠利, Kundjanasith Thonglek, 市川昊平, 高橋慧智

信学技報　122　(201)　21-24　2022年10月
機械学習に基づくジョブスケジューリングのためのGANによるデータ拡張

石井翔, 高橋慧智, 下村陽一, 滝沢寛之

研究報告ハイパフォーマンスコンピューティング（HPC）　2022-HPC-185　(15)　1-8　2022年7月
計算特性に着目した実行時間予測に基づくリアルタイム洪水シミュレーションの動的資源割当

青柳嶺, 高橋慧智, 下村陽一, 滝沢寛之

研究報告ハイパフォーマンスコンピューティング（HPC）　2022-HPC-185　(27)　1-9　2022年7月
Opimon: A Transparent, Low-overhead Monitoring System for OpenFlow Networks 査読有り

Wassapon WATANAKEESUNTORN, Keichi TAKAHASHI, Chawanat NAKASAN, Kohei ICHIKAWA, Hajimu IIDA

IEICE Transactions on Communications　E105-B　(4)　485-493　2022年4月
出版者・発行元：Institute of Electronics, Information and Communications Engineers (IEICE)
DOI： 10.1587/transcom.2021ebp3083 　

ISSN：0916-8516

eISSN：1745-1345

詳細を見る詳細を閉じる

OpenFlow is a widely adopted implementation of the Software-Defined Networking (SDN) architecture. Since conventional network monitoring systems are unable to cope with OpenFlow networks, researchers have developed various monitoring systems tailored for Open- Flow networks. However, these existing systems either rely on a specific controller framework or an API, both of which are not part of the Open- Flow specification, and thus limit their applicability. This article proposes a transparent and low-overhead monitoring system for OpenFlow networks, referred to as Opimon. Opimon monitors the network topology, switch statistics, and flow tables in an OpenFlow network and visualizes the result through a web interface in real-time. Opimon monitors a network by interposing a proxy between the controller and switches and intercepting every OpenFlow message exchanged. This design allows Opimon to be compatible with any OpenFlow switch or controller. We tested the functionalities of Opimon on a virtual network built using Mininet and a large-scale international OpenFlow testbed (PRAGMA-ENT). Furthermore, we measured the performance overhead incurred by Opimon and demonstrated that the overhead in terms of latency and throughput was less than 3% and 5%, respectively.
A Cost Model for Compilers Based on Transfer Learning. 査読有り

Yuta Sasaki, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

IPDPS Workshops　942-951　2022年

DOI： 10.1109/IPDPSW55747.2022.00152 　
Automated selection of build configuration based on machine learning. 査読有り

Reo Furuhata, Minglu Zhao, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

IPDPS Workshops　934-941　2022年

DOI： 10.1109/IPDPSW55747.2022.00151 　
Sparse Communication for Federated Learning. 査読有り

Kundjanasith Thonglek, Keichi Takahashi, Kohei Ichikawa, Chawanat Nakasan, Pattara Leelaprute, Hajimu Iida

ICFEC　1-8　2022年

DOI： 10.1109/ICFEC54809.2022.00008 　
Automated Quantization and Retraining for Neural Network Models Without Labeled Data. 査読有り

Kundjanasith Thonglek, Keichi Takahashi, Kohei Ichikawa, Chawanat Nakasan, Hidemoto Nakada, Ryousei Takano, Pattara Leelaprute, Hajimu Iida

IEEE Access　10　73818-73834　2022年

DOI： 10.1109/ACCESS.2022.3190627 　
Acar: An application-aware network routing system using SRv6. 査読有り

Tomoki Sugiura, Keichi Takahashi, Kohei Ichikawa, Hajimu Iida

CCNC　751-752　2022年

DOI： 10.1109/CCNC49033.2022.9700608 　
Difficulty of detecting overstated dataset size in Federated Learning

研究報告マルチメディア通信と分散処理（DPS）　2021-DPS-189　(10)　2021年12月
Enhancing the Prediction Accuracy of Solar Power Generation using a Generative Adversarial Network 査読有り

Kundjanasith Thonglek, Kohei Ichikawa, Keichi Takahashi, Chawanat Nakasan, Kazufumi Yuasa, Tadatoshi Babasaki, Hajimu Iida

IEEE Green Energy and Smart Systems Conference (IGESSC)　2021年11月
出版者・発行元：IEEE
DOI： 10.1109/igessc53124.2021.9618702 　

詳細を見る詳細を閉じる

Solar power is the most widely used green energy. However, using solar power generation as a stable power supply remains challenging since the power output is difficult to predict. Accurate prediction of solar power generation enables efficient control of the amount of stored electricity in batteries to produce a stable supply of electricity. This paper aims to build a highly accurate solar power prediction model. For this purpose, we design a neural network model based on Long Short-Term Memory (LSTM) to predict the future solar power generation using past solar power generation and weather forecasts. Since a large and diverse dataset is required to train an accurate prediction model, we develop a neural network based on Generative Adversarial Network (GAN) to generate artificial datasets from the original training dataset to increase the amount and diversity of the training dataset. Additionally, stratified k-fold cross-validation is used to eliminate learning deviation during training. As a result, the proposed neural network model based on GAN improved the R2 score of LSTM from 0.750 to 0.805 with stratified k-fold cross-validation.
kEDM: A Performance-portable Implementation of Empirical Dynamic Modeling using Kokkos 査読有り

Keichi Takahashi, Wassapon Watanakeesuntorn, Kohei Ichikawa, Joseph Park, Ryousei Takano, Jason Haga, George Sugihara, Gerald M. Pao

Practice & Experience in Advanced Research Computing (PEARC 2021)　8-8　2021年7月
出版者・発行元：ACM
DOI： 10.1145/3437359.3465571 　
SRv6を用いたアプリケーションの特性を考慮した通信経路制御手法

杉浦智基, 高橋慧智, 市川昊平, 飯田元

研究報告インターネットと運用技術（IOT）　2021-IOT-53　(18)　2021年5月
Comparative Performance Study of Lightweight Hypervisors Used in Container Environment 査読有り

215-223　2021年4月
出版者・発行元：None
DOI： 10.5220/0010440502150223 　
A codesign framework for online data analysis and reduction 査読有り

Kshitij Mehta, Bryce Allen, Matthew Wolf, Jeremy Logan, Eric Suchyta, Swati Singhal, Jong Y. Choi, Keichi Takahashi, Kevin Huck, Igor Yakushin, Alan Sussman, Todd Munson, Ian Foster, Scott Klasky

Concurrency and Computation: Practice and Experience　2021年

DOI： 10.1002/cpe.6519 　

ISSN：1532-0626

eISSN：1532-0634

詳細を見る詳細を閉じる

Science applications preparing for the exascale era are increasingly exploring in situ computations comprising of simulation-analysis-reduction pipelines coupled in-memory. Efficient composition and execution of such complex pipelines for a target platform is a codesign process that evaluates the impact and tradeoffs of various application- and system-specific parameters. In this article, we describe a toolset for automating performance studies of composed HPC applications that perform online data reduction and analysis. We describe Cheetah, a new framework for composing parametric studies on coupled applications, and Savanna, a runtime engine for orchestrating and executing campaigns of codesign experiments. This toolset facilitates understanding the impact of various factors such as process placement, synchronicity of algorithms, and storage versus compute requirements for online analysis of large data. Ultimately, we aim to create a catalog of performance results that can help scientists understand tradeoffs when designing next-generation simulations that make use of online processing techniques. We illustrate the design of Cheetah and Savanna, and present application examples that use this framework to conduct codesign studies on small clusters as well as leadership class supercomputers.
Massively Parallel Causal Inference of Whole Brain Dynamics at Single Neuron Resolution 査読有り

Wassapon Watanakeesuntorn, Keichi Takahashi, Kohei Ichikawa, Joseph Park, George Sugihara, Ryousei Takano, Jason Haga, Gerald M. Pao

2020 IEEE 26th International Conference on Parallel and Distributed Systems (ICPADS)　196-205　2020年12月
出版者・発行元：IEEE
DOI： 10.1109/icpads51040.2020.00035 　

ISSN：1521-9097

詳細を見る詳細を閉じる

Empirical Dynamic Modeling (EDM) is a nonlinear time series causal inference framework. The latest implementation of EDM, cppEDM, has only been used for small datasets due to computational cost. With the growth of data collection capabilities, there is a great need to identify causal relationships in large datasets. We present mpEDM, a parallel distributed implementation of EDM optimized for modern GPU-centric supercomputers. We improve the original algorithm to reduce redundant computation and optimize the implementation to fully utilize hardware resources such as GPUs and SIMD units. As a use case, we run mpEDM on AI Bridging Cloud Infrastructure (ABCI) using datasets of an entire animal brain sampled at single neuron resolution to identify dynamical causation patterns across the brain. mpEDM is 1,530x faster than cppEDM and a dataset containing 101,729 neuron was analyzed in 199 seconds on 512 nodes. This is the largest EDM causal inference achieved to date.
Federated Learning of Neural Network Models with Heterogeneous Structures 査読有り

Kundjanasith Thonglek, Keichi Takahashi, Kohei Ichikawa, Hajimu Iida, Chawanat Nakasan

2020 19th IEEE International Conference on Machine Learning and Applications (ICMLA)　735-740　2020年12月
出版者・発行元：IEEE
DOI： 10.1109/icmla51294.2020.00120 　

詳細を見る詳細を閉じる

Federated learning trains a model on a centralized server using datasets distributed over a large number of edge devices. Applying federated learning ensures data privacy because it does not transfer local data from edge devices to the server. Existing federated learning algorithms assume that all deployed models share the same structure. However, it is often infeasible to distribute the same model to every edge device because of hardware limitations such as computing performance and storage space. This paper proposes a novel federated learning algorithm to aggregate information from multiple heterogeneous models. The proposed method uses weighted average ensemble to combine the outputs from each model. The weight for the ensemble is optimized using black box optimization methods. We evaluated the proposed method using diverse models and datasets and found that it can achieve comparable performance to conventional training using centralized datasets. Furthermore, we compared six different optimization methods to tune the weights for the weighted average ensemble and found that tree parzen estimator achieves the highest accuracy among the alternatives.
共有型IoT資源利用アプリケーションのためのデータフロープログラミング

村木暢哉, 木戸善之, 高橋慧智, 山田拓哉, 伊達進, 梅谷麗, 石橋靖嗣, 下條真司

研究報告システムソフトウェアとオペレーティング・システム（OS）　2020-OS-150　(5)　2020年7月
Retraining Quantized Neural Network Models with Unlabeled Data 査読有り

Kundjanasith Thonglek, Keichi Takahashi, Kohei Ichikawa, Chawanat Nakasan, Hidemoto Nakada, Ryousei Takano, Hajimu Iida

2020 International Joint Conference on Neural Networks (IJCNN)　2020年7月
出版者・発行元：IEEE
DOI： 10.1109/ijcnn48605.2020.9207190 　

詳細を見る詳細を閉じる

Running neural network models on edge devices is attracting much attention by neural network researchers since edge computing technology is becoming more powerful than ever. However, deploying large neural network models on edge devices is challenging due to the limitation in available computing resources and storage space. Therefore, model compression techniques have been recently studied to reduce the model size and fit models on resource-limited edge devices. Compressing neural network models reduces the size of a model, but also degrades the accuracy of the model since it reduces the precision of weights in the model. Consequently, a retraining method is required to recover the accuracy of compressed models. Most existing retraining methods require the original labeled training datasets to retrain the models, but labeling is a time-consuming process. In particular, we cannot always access the original labeled datasets because of privacy policies and license limitations. In this paper, we propose a method to retrain a compressed neural network model with an unlabeled dataset that is different from the original labeled dataset. We compress the neural network model using quantization to decrease the size of the model. Subsequently, the compressed model is retrained by our proposed retraining method without using a labeled dataset to recover the accuracy of the model. We compared the proposed retraining method against the conventional retraining. The proposed method reduced the size of VGG-16 and ResNet-50 by 81.10% and 52.45%, respectively without significant accuracy loss. In addition, our proposed retraining method is clearly faster than the conventional retraining method.
ADIOS 2: The Adaptable Input Output System. A framework for high-performance data management 査読有り

William F. Godoy, Norbert Podhorszki, Ruonan Wang, Chuck Atkins, Greg Eisenhauer, Junmin Gu, Philip Davis, Jong Choi, Kai Germaschewski, Kevin Huck, Axel Huebl, Mark Kim, James Kress, Tahsin Kurc, Qing Liu, Jeremy Logan, Kshitij Mehta, George Ostrouchov, Manish Parashar, Franz Poeschel, David Pugmire, Eric Suchyta, Keichi Takahashi, Nick Thompson, Seiji Tsutsumi, Lipeng Wan, Matthew Wolf, Kesheng Wu, Scott Klasky

SoftwareX　12　100561-100561　2020年7月
出版者・発行元：Elsevier {BV}
DOI： 10.1016/j.softx.2020.100561 　

ISSN：2352-7110
Massively Parallel Causal Inference of Whole Brain Dynamics at Single Neuron Resolution.

Wassapon Watanakeesuntorn, Keichi Takahashi, Kohei Ichikawa, Joseph Park, George Sugihara, Ryousei Takano, Jason Haga, Gerald M. Pao

CoRR　abs/2011.11082　2020年
Integrating SDN-Enhanced MPI with Job Scheduler to Support Shared Clusters 招待有り

Keichi Takahashi, Susumu Date, Yasuhiro Watashiba, Yoshiyuki Kido, Shinji Shimojo

Sustained Simulation Performance 2018 and 2019　149-159　2020年
出版者・発行元：Springer International Publishing
DOI： 10.1007/978-3-030-39181-2_13 　
In Situ/In Transitアプローチを用いた大規模数値解析におけるポスト処理効率化

堤誠司, 藤田直行, 伊藤浩之, 大日向大地, 井上敬介, 松村洋祐, 高橋慧智, Greg Eisenhauer, Norbert Podhorszki, Scott Klasky

第33回数値流体力学シンポジウム　2019年11月
Improving Resource Utilization in Data Centers using an LSTM-based Prediction Model. 査読有り

Kundjanasith Thonglek, Kohei Ichikawa, Keichi Takahashi, Hajimu Iida, Chawanat Nakasan

Workshop on Monitoring and Analysis for High Performance Computing Systems Plus Applications, HPCMASPA 2019　1-8　2019年
出版者・発行元：IEEE
DOI： 10.1109/CLUSTER.2019.8891022 　
A Codesign Framework for Online Data Analysis and Reduction. 査読有り

Kshitij Mehta, Ian T. Foster, Scott Klasky, Bryce Allen, Matthew Wolf, Jeremy Logan, Eric Suchyta, Jong Choi, Keichi Takahashi, Igor Yakushin, Todd Munson

Workflows in Support of Large-Scale Science, WORKS 2019　11-20　2019年
出版者・発行元：IEEE
DOI： 10.1109/WORKS49585.2019.00007 　
Connected-HPCに向けたネットワークの動的管理技術の設計と実装

森本弘明, 高橋慧智, 山田拓哉, 木戸善之, 伊達進, 下條真司

日本ソフトウェア科学会第16回ディペンダブルシステムワークショップ (DSW2018)　2018年12月
医療応用を考慮した動的構成変更可能計算機クラスタの検討

三澤明寛, 高橋慧智, 渡場康弘, 伊達進, 吉川隆士, 阿部洋丈, 野崎一徳, 木戸善之, Lee CHONHO, 下條真司

日本ソフトウェア科学会第16回ディペンダブルシステムワークショップ (DSW2018)　2018年12月
Towards Connected-HPC

Hiroaki Morimoto, Takuya Yamada, Keichi Takahashi, Yoshiyuki Kido, Susumu Date, Shinji Shimojo

SEAIP2018　2018年11月
A Traffic Simulator with Intra-node Parallelism for Designing High-performance Interconnects. 査読有り

Yohei Takigawa, Keichi Takahashi, Susumu Date, Yoshiyuki Kido, Shinji Shimojo

2018 International Conference on High Performance Computing & Simulation, HPCS 2018, Orleans, France, July 16-20, 2018　445-451　2018年
出版者・発行元：IEEE
DOI： 10.1109/HPCS.2018.00077 　
UnisonFlow: A Software-Defined Coordination Mechanism for Message-Passing Communication and Computation. 査読有り

Keichi Takahashi, Susumu Date, Dashdavaa Khureltulga, Yoshiyuki Kido, Hiroaki Yamanaka, Eiji Kawai, Shinji Shimojo

IEEE Access　6　23372-23382　2018年
出版者・発行元：Institute of Electrical and Electronics Engineers Inc.
DOI： 10.1109/ACCESS.2018.2829532 　

ISSN：2169-3536

詳細を見る詳細を閉じる

Message passing interface (MPI) communication performance is becoming one of the key factors heavily affecting the total performance of data-intensive applications running on computer clusters. Our software-defined networking (SDN)-enhanced MPI improves the performance of communication over interconnects by integrating flexible and dynamic network controllability of SDN into MPI. We have demonstrated that the acceleration of individual MPI communication primitives is feasible through our past work on the SDN-enhanced MPI. However, real-world MPI applications have not benefited from such accelerated communication primitives through our research achievements to date, because each of the distinct network control algorithms designed for various MPI communication primitives cannot be activated and coordinated with the execution of the MPI application. Therefore, this paper proposes UnisonFlow, a software-defined coordination mechanism for the SDN-enhanced MPI that performs network control in synchronization with the execution of applications. An experiment conducted on a real-computer cluster verifies that the interconnect control can be successfully performed in synchronization with the execution of the application. Furthermore, the synchronization is performed with a low overhead and its performance penalty is practically negligible.
SDNを用いた高機能なインターコネクトの実現

森本弘明, Dashdavaa Khureltulga, 高橋慧智, 木戸善之, 伊達進, 下條真司

第15回ディペンダブルシステムワークショップ (DSW 2017)　2017年12月
Design and Implementation of SDN-enhanced MPI Broadcast Targeting a Fat-Tree Interconnect. 査読有り

Hiroaki Morimoto, Khureltulga Dashdavaa, Keichi Takahashi, Yoshiyuki Kido, Susumu Date, Shinji Shimojo

2017 International Conference on High Performance Computing & Simulation (HPCS)　252-258　2017年
出版者・発行元：IEEE
DOI： 10.1109/HPCS.2017.46 　
PFAnalyzer: A Toolset for Analyzing Application-Aware Dynamic Interconnects. 査読有り

Keichi Takahashi, Susumu Date, Dashdavaa Khureltulga, Yoshiyuki Kido, Shinji Shimojo

2017 IEEE INTERNATIONAL CONFERENCE ON CLUSTER COMPUTING (CLUSTER)　789-796　2017年
出版者・発行元：IEEE
DOI： 10.1109/CLUSTER.2017.18 　

ISSN：1552-5244

詳細を見る詳細を閉じる

Recent rapid scale out of high performance computing systems has rapidly and continuously increased the scale and complexity of the interconnects. As a result, current static and over-provisioned interconnects are becoming cost-ineffective. Against this background, we have been working on the integration of network programmability into the interconnect control, based on the idea that dynamically controlling the packet flow in the interconnect according to the communication pattern of applications can increase the utilization of interconnects and improve application performance. Interconnect simulators come in handy especially when investigating the performance characteristics of interconnects with different topologies and parameters. However, little effort has been put towards the simulation of packet flow in dynamically controlled interconnects, while simulators for static interconnects have been extensively researched and developed. To facilitate analysis on the performance characteristics of dynamic interconnects, we have developed PFAnalyzer. PFAnalyzer is a toolset composed of PFSim, an interconnect simulator specialized for dynamic interconnects, and PFProf, a profiler. PFSim allows interconnect researchers and designers to investigate congestion in the interconnect for an arbitrary cluster configuration and a set of communication patterns collected by PFProf. PFAnalyzer is used to demonstrate how dynamically controlling the interconnects can reduce congestion and potentially improve the performance of applications.
Dynamic Reconfiguration of Computer Platforms at the Hardware Device Level for High Performance Computing Infrastructure as a Service. 査読有り

Akihiro Misawa, Susumu Date, Keichi Takahashi, Takashi Yoshikawa, Masahiko Takahashi, Masaki Kan, Yasuhiro Watashiba, Yoshiyuki Kido, Chonho Lee, Shinji Shimojo

Cloud Computing and Service Science - 7th International Conference, CLOSER 2017, Porto, Portugal, April 24-26, 2017, Revised Selected Papers　177-199　2017年
出版者・発行元：Springer
DOI： 10.1007/978-3-319-94959-8_10 　
Highly Reconfigurable Computing Platform for High Performance Computing Infrastructure as a Service: Hi-IaaS. 査読有り

Akihiro Misawa, Susumu Date, Keichi Takahashi, Takashi Yoshikawa, Masahiko Takahashi, Masaki Kan, Yasuhiro Watashiba, Yoshiyuki Kido, Chonho Lee, Shinji Shimojo

CLOSER 2017 - Proceedings of the 7th International Conference on Cloud Computing and Services Science, Porto, Portugal, April 24-26, 2017.　135-146　2017年
出版者・発行元：SciTePress
DOI： 10.5220/0006302501350146 　
MPI通信パターンに基づくSDN制御を高速化するカーネルモジュールの試作と評価

高橋慧智, Khureltulga Dashdavaa, 木戸善之, 伊達進, 下條真司

情報処理学会研究報告　2016-OS-137　(13)　2016年5月
Network Access Control Towards Fully-Controlled Cloud Infrastructure. 査読有り

Takuya Yamada, Keichi Takahashi, Masaya Muraki, Susumu Date, Shinji Shimojo

2016 8TH IEEE INTERNATIONAL CONFERENCE ON CLOUD COMPUTING TECHNOLOGY AND SCIENCE (CLOUDCOM 2016)　452-455　2016年
出版者・発行元：IEEE
DOI： 10.1109/CloudCom.2016.0076 　

ISSN：2330-2194

詳細を見る詳細を閉じる

Recently, researchers' and scientists' interest and concern to Internet of Things (IoT) have been remarkably increasing. A diversity of IoT devices such as mobile phones, sensors and even scientific measurement facilities have been connected to the Internet and then generating an enormous amount of data. From the demands on computational resources enough to analyze such data, the utilization of the cloud has been a major trend in these days. Taking aggregation and distribution of data from and to IoT devices on the cloud into consideration, however, access control to such data gives rise to an important problem. Each of IoT devices may have a security policy and each user may have a different attribute. For achieving safe access control to data, a fully-controlled infrastructure where access to network resources is controlled as well as computational resources is required. From such a consideration, this paper proposes an access-controlled networking mechanism that dynamically organizes a flexible and secure network linking IoT devices, computational resources and users on the cloud, based on user's attribute and IoT device security policies. The architecture of FlowSieve, which we have designed and implemented in this preliminary stage of the research, is presented as well as our envisaged fully access-controlled cloud for secure data access.
Toward Flexible Supercomputing and Visualization System 招待有り

Keichi Takahashi

Sustained Simulation Performance 2015　77-93　2015年

DOI： 10.1007/978-3-319-20340-9_7 　
Design and implementation of control sequence generator for SDN-enhanced MPI. 査読有り

Baatarsuren Munkhdorj, Keichi Takahashi, Dashdavaa Khureltulga, Yasuhiro Watashiba, Yoshiyuki Kido, Susumu Date, Shinji Shimojo

Proceedings of NDM 2015: 5th International Workshop on Network-Aware Data Management - Held in conjunction with SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis　4-9　2015年
出版者・発行元：Association for Computing Machinery, Inc
DOI： 10.1145/2832099.2832103 　

詳細を見る詳細を閉じる

MPI (Message Passing Interface) offers a suite of APIs for inter-process communication among parallel processes. We have approached to the acceleration of MPI collective communication such as MPI Bcast and MPI Allreduce, taking advantage of network programmability brought by Software Definned Networking (SDN). The basic idea is to allow a SDN controller to dynamically control the packet ows generated by MPI collective communication based on the communication pattern and the underlying network conditions. Al- Though our research have succeeded to accelerate an MPI collective communication in terms of execution time, the switching of network control functionality for MPI collec- Tive communication along MPI program execution have not been considered yet. This paper presents a mechanism that provides the control sequence for SDN controller to control packet ows based on the communication plan for the entire MPI application. The control sequence encloses a chronologically ordered list of the MPI collectives operated in the MPI application and the process-related information of each in the list. To verify if the SDN-enhanced MPI collectives can be used in combination with the proposed mechanism, the envisioned environment was prototyped. As a result, SDN-enhanced MPI collectives were able to be used in com-bination.
MPI_Reduce algorithm for OpenFlow-enabled network. 査読有り

Pisit Makpaisit, Kohei Ichikawa, Putchong Uthayopas, Susumu Date, Keichi Takahashi, Dashdavaa Khureltulga

2015 15TH INTERNATIONAL SYMPOSIUM ON COMMUNICATIONS AND INFORMATION TECHNOLOGIES (ISCIT)　261-264　2015年
出版者・発行元：IEEE
DOI： 10.1109/ISCIT.2015.7458357 　

詳細を見る詳細を閉じる

The MPI reduction operation such as MPI_Reduce and MPI_Allreduce are frequently used and time-consuming operations. The performance enhancement of these operations can substantially speed up large-scale parallel applications. In this paper, a greedy based MPI_Reduce algorithm called Greedy Shortest Binomial Tree (GSBT) is proposed. This proposed algorithm leverages SDN technology and OpenFlow network to speed up MPI reduction operations. This is accomplished using network topology information from the OpenFlow controller to reduce overall hops in message transmission. The implementation of the proposed algorithm by modifying MPI library and OpenFlow controller is presented. The proposed GSBT algorithm has been evaluated in a real test-bed to compare with the traditional approaches used in both MPICH and Open MPI. The result shows that GSBT algorithm is faster than standard algorithms 30.48-66.35% for Open MPI and faster 50.77-82.89% for MPICH when message size between 2 KB - 24 KB.
An Empirical Study of SDN-accelerated HPC Infrastructure for Scientific Research. 査読有り

Susumu Date, Hirotake Abe, Dashdavaa Khureltulga, Keichi Takahashi, Yoshiyuki Kido, Yasuhiro Watashiba, Pongsakorn U.-Chupala, Kohei Ichikawa, Hiroaki Yamanaka, Eiji Kawai, Shinji Shimojo

2015 INTERNATIONAL CONFERENCE ON CLOUD COMPUTING RESEARCH AND INNOVATION (ICCCRI)　89-96　2015年
出版者・発行元：IEEE
DOI： 10.1109/ICCCRI.2015.13 　

詳細を見る詳細を閉じる

High performance computing is required for Big Science application because the proliferation and huge amount of scientific data that needs to be analyzed is a serious problem. Traditionally, network resources were generally assumed as a static resource users cannot control on demand. By integrating network programmability to every stage of a scientific workflow, this study explores a next-generation high performance computing infrastructure where both computational and network resources are flexibly sliced and efficiently leveraged based on the resource requirements of the scientific applications. Technically, Software Defined Networking has been adopted as a key technology for this purpose. In this paper the concept and goals of a next-generation high performance computing infrastructure is introduced and the current status of our research is discussed.
Concept and Design of SDN-Enhanced MPI Framework. 査読有り

Keichi Takahashi, Dashdavaa Khureltulga, Baatarsuren Munkhdorj, Yoshiyuki Kido, Susumu Date, Hiroaki Yamanaka, Eiji Kawai, Shinji Shimojo

2015 FOURTH EUROPEAN WORKSHOP ON SOFTWARE DEFINED NETWORKS - EWSDN 2015　109-110　2015年
出版者・発行元：IEEE
DOI： 10.1109/EWSDN.2015.72 　

詳細を見る詳細を閉じる

In general, modern high-performance computing systems are built as cluster systems. We have been investigating the feasibility of optimizing MPI communications by integrating the dynamic network control realized by SDN. In this paper, we present a concept of a generic SDN enhanced MPI framework; an application-aware network control mechanism specifically for MPI applications.
Performance evaluation of SDN-enhanced MPI allreduce on a cluster system with fat-tree interconnect. 査読有り

Keichi Takahashi, Dashdavaa Khureltulga, Yasuhiro Watashiba, Yoshiyuki Kido, Susumu Date, Shinji Shimojo

2014 INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING & SIMULATION (HPCS)　784-792　2014年
出版者・発行元：IEEE
DOI： 10.1109/HPCSim.2014.6903768 　

詳細を見る詳細を閉じる

Nowadays, supercomputers play an essential role in high-performance computing. In general, modern supercomuputers are built as a cluster system, which is a system of multiple computers interconnected on a network. In coding a parallel program on such a cluster system, MPI (Message Passing Interface) is utilized. In this paper, we aim to reduce the execution time of MPI_Allreduce, a frequently used MPI collective communication in many simulation codes. To this end, we have integrated network programmability by Software Defined Networking into MPI_Allreduce so that it effectively uses the bandwidth of the interconnect of the cluster system. An experiment conducted on a cluster system with fat-tree interconnect indicates that our proposed MPI_Allreduce is superior to MPI_Allreduce in OpenMPI implementations.

︎全件表示 ︎最初の5件までを表示

講演・口頭発表等 12

Clustering Based Job Runtime Prediction for Backfilling Using Classification

Hang Cui, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

HPC Asia 2024
Using lossy compression for interactive analysis over network

Rei Aoyagi, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

HPC Asia 2024
Calibrating Simulations of Quantum Annealers for Predictive Models

Michael Zielewski, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

HPC Asia 2024
A Power Management Method to Improve Energy Budget Utilization

Sho Ishii, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

HPC Asia 2024
A Proposal of Automatic Parallelization using Transformer-based Large Language Models

Soratouch Pornmaneerattanatri, Keichi Takahashi, Yutaro Kashiwa, Kohei Ichikawa, Hajimu Iida

HPC Asia 2024
An energy-aware job scheduling method supporting on-demand job execution

Daiki Nakai, Keichi Takahashi, Yoichi Shimomura, Hiroyuki Takizawa

HPC Asia 2024
Implementation and Application of High-Performance Empirical Dynamic Modeling

高橋慧智

学際大規模情報基盤共同利用・共同研究拠点第15回シンポジウム　2023年7月7日
Challenges in Scaling Empirical Dynamic Modeling 招待有り

Keichi Takahashi

34th Workshop on Sustained Simulation Performance (WSSP34)　2022年10月25日
Accelerating Empirical Dynamic Modeling using High Performance Computing 招待有り

Keichi Takahashi

University of Melbourne Quantitative Research Methods Network (QMNET)　2021年8月20日
An MPI Framework for HPC Clusters Deployed with Software-Defined Networking 国際会議招待有り

高橋慧智

27th Workshop on Sustained Simulation Performance (WSSP27)　2018年3月
Towards Realizing a Dynamic and MPI Application-aware Interconnect with SDN 国際会議招待有り

高橋慧智

26th Workshop on Sustained Simulation Performance (WSSP26)　2017年10月
Control Sequence Generator for Generic SDN-enhanced MPI Framework 国際会議

高橋慧智

28th Pacific Rim Application and Grid Middleware Assembly (PRAGMA 28)　2015年4月

︎全件表示 ︎最初の5件までを表示

共同研究・競争的資金等の研究課題 6

性能最適化の知見を自ら発掘する自動チューニング技術の開発

高橋慧智

2023年4月1日～ 2026年3月31日
NISQ時代を見据えたバッチ型量子回路シミュレータの開発

2024年4月～ 2025年3月
Accelerating Manifold Inference from Neural Dynamics (MIND) using HPC

提供機関：Okinawa Institute of Science and Technology

制度名：SHINKA Grant 2023

2023年6月～ 2024年3月
多階層マルチスケール・ソフトウェア分析基盤に関する研究開発

飯田元, 藤原賢二, 崔恩瀞

提供機関：Japan Society for the Promotion of Science

制度名：Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (B)

研究種目：Grant-in-Aid for Scientific Research (B)

研究機関：Nara Institute of Science and Technology

2021年4月～ 2024年3月
Implementation and Application of High-Performance Empirical Dynamic Modeling

高橋慧智, Gerald Pao

2022年4月～ 2023年3月
In-situワークフローのための適応的な計算資源配分フレームワーク

高橋慧智

提供機関：Japan Society for the Promotion of Science

制度名：Grants-in-Aid for Scientific Research Grant-in-Aid for Early-Career Scientists

研究種目：Grant-in-Aid for Early-Career Scientists

研究機関：Nara Institute of Science and Technology

2020年4月～ 2023年3月

詳細を見る詳細を閉じる

高性能計算システムにおけるストレージボトルネックを回避するため，ストレージを介さず，シミュレーションや可視化アプリケーション間で直接データを授受するIn-situ処理が注目されている．In-situ処理のスループットおよび資源利用率の最大化のためには，各アプリケーション間でスループットが均等になるように計算資源を配分することが不可欠である．本研究では，In-situ処理において各アプリケーションへ計算資源を自動的に配分するフレームワークの構築を目指している．今年度は研究計画にしたがい，「課題2. 計算資源配分の決定」および「課題3. ワークフローの再構成」に取り組んだ．まず計算資源配分をアプリケーション間で動的に配分するアルゴリズムの開発のため，モデルとなるIn-situワークフローを開発した．具体的には，3次元反応拡散系のシミュレーション，および，シミュレーション結果をリアルタイムに可視化する複数のアプリケーション (等値曲面の抽出・可視化, 断面の可視化, 要約統計量の算出など) を開発し，昨年度開発した性能計測システムを用いて様々な資源配分の下で各アプリケーションの性能情報を収集した．得られた性能情報を分析するとともに，オートスケーリングやストリーミング処理に関する研究の文献調査を実施し，計算資源配分アルゴリズムを設計した．また，計算資源配分を動的に変更するためのワークフロー再構成手法を検討し，実現に向けた課題を整理した．

︎全件表示 ︎最初の5件までを表示

担当経験のある科目(授業) 6

コンピュータ実習I　東北大学
コンピュータ実習II　東北大学
ソフトウェアシステム構築論　奈良先端科学技術大学院大学
マルチメディア情報処理　奈良工業高等専門学校
情報科学特別講義　奈良先端科学技術大学院大学
情報工学実験III　奈良工業高等専門学校

︎全件表示 ︎最初の5件までを表示