ACSI2016 プログラム
基調講演・招待講演・チュートリアル情報
- 基調講演
- 1/18(月) 14:00-15:30：
九州大学
安浦 寛人 先生
「都市OSの構築に向けて」
情報通信技術は、計算／通信速度やデータ量を飛躍的に増大させ、人類が築いてきた文化や文明を本質的に変化させようとしている。我々が生活する都市の基盤も情報通信技術に益々依存し、新しい都市計画のあり方が模索されている。環境問題や人口問題等あらゆる都市の問題の解決のために、都市OSを構築する必要がある。都市OSの基本的な構造と必要となる技術を研究している「九州大学共進化社会システム拠点」の活動も紹介する。
- 1/18(月) 14:00-15:30：
- 招待講演
- 1/19(火) 13:00-14:30：
米国サンディア国立研究所
Michael Heroux 氏
"Strategies for Next Generation HPC Applications and Systems"
The extreme-scale computing community is several years into a highly disruptive period of change. New commodity performance curves must be incorporated into application designs, and the orders of magnitude in performance potential will increase the demand to couple physics and scales into a single integrated execution environment.
In this talk we discuss several aspects of next-generation scientific and engineering applications that will be particularly challenging and easy as we prepared for new systems. We touch on application design, parallel programming strategies, software distribution and resiliency, and introduce how a productivity focus can provide new approaches to better applications.
- 1/19(火) 13:00-14:30：
- チュートリアル
- チュートリアル1
1/19(火) 9:15-10:45：
九州大学
藤澤 克樹 先生
「高速かつ省電力なグラフ解析とその実応用」
新しいスーパーコンピュータの応用として大規模なグラフ解析が注目を集めている. グラフ解析の応用分野としては大規模災害等での避難誘導計画，社会公共政策や企業経営等のためソーシャル・ネットワーク等の大規模データの有効活用等が想定されているが, 非常に計算量やデータ量さらに電力使用量などの規模が大きく従来の手法では処理が困難である. そのためハイパフォーマンスコンピューティング分野の技術を用いて大規模なグラフ解析を行う研究がさかんに行われており, その中からGraph500 ベンチマークを題材に高速かつ省電力なグラフ解析ソフトウェアの開発方法について解説を行う. 現在、サイバーフィジカルシステムにおいて、ヒト・モノのモビリティに関する数理モデル及び最適化手法としてもグラフ解析の利用が開始されている。このような最先端理論（Algorithm Theory）+ 大規模実データ（BigData）+ 最新計算技術（Computation）の有機的な組合せによる最新のグラフ解析の応用についても紹介を行う．
- チュートリアル2
1/20(水) 13:45-15:15：
東北大学
岡谷 貴之 先生
「深層学習とその画像認識への応用に関する最近の研究動向」
画像認識は深層学習がこれまで最も成功した問題であると言える．画像に写る物体の種類を認識する問題を筆頭に，画像認識の数々のタスクにおいて，従来法に比べ圧倒的に高い認識性能を実現している．これまで難しかった画素単位での物の認識（セグメンテーション）や，画像のコンテンツを文章で記述する問題も，今や現実的になりつつある．この他にも深層学習を使った画像の生成のように，応用の拡大も早いペースで進んでいる．本チュートリアルでは，深層学習の基礎から始め，応用を中心とした最近の研究の進展を解説する．
- チュートリアル1
企業・団体展示（中ホール１）
1/18(月) 13:30--18:00，1/19(火) 9:15--20:00の間，常時展示．
展示紹介のためのインデキシング を2日目午後に予定しております．
参加企業・団体一覧（敬称略）
詳細はACSIS2016 企業・団体展示参照．
- アマゾン ウェブ サービス ジャパン株式会社
- 株式会社データダイレクト・ネットワークス・ジャパン
- Allinea Software株式会社
- 日本電気株式会社
- 株式会社 日立製作所
- 富士通株式会社
スポンサード・ランチ・セッション
1/19(火) 12:00--13:00．
2日目の昼に，アマゾン ウェブ サービス ジャパン株式会社によるスポンサード・ランチ・セッションを予定しております．
セッション構成
プログラム
発表時間・・・[long]: 30min, [short]: 20min, [oversea]: 30min
13:30∼14:00
- オープニング (Chair: Susumu Date (Osaka University))
- 開会の辞
- 一般論文表彰
- 開会の辞
14:00∼15:30
- 基調講演 (Chair: Masahiro Yasugi (Kyushu Institute of Technology))
- 「都市OSの構築に向けて」
安浦 寛人 (九州大学)
- 「都市OSの構築に向けて」
16:00∼18:00
- [1A] numerical analysis and performance tuning (Chair: Maho Nakata (RIKEN))
- (1) Variable Preconditioning of Krylov Subspace Methods for Hierarchical Matrices with Adaptive Cross Approximation [oversea]
Akihiro Ida (Kyoto University), Takeshi Iwashita (Hokkaido University), Takeshi Mifune (Kyoto University) and Yasuhito Takahashi (Doshisha University)
出典20th International Conference on the Computation of Electromagnetic Fields (Compumag 2015)
- (2) Parallel Iterative Solvers based on Domain Decomposition with Robust Preconditioners for Ill-conditioned Problems [long]
Kengo Nakajima (The University of Tokyo)
- (3) Towards Code Selection by Auto-tuning with ppOpen-AT: An Example of Computation Kernels from FDM [long]
Takahiro Katagiri, Masaharu Matsumoto and Satoshi Ohshima (The University of Tokyo/JST CREST)
- ACSI2016 Outstanding Research Award
[28] Accelerating Tomographic Reconstruction With Efficient Interpolation [long]
Hiroshi Inoue (IBM Research - Tokyo/The University of Tokyo)
- (1) Variable Preconditioning of Krylov Subspace Methods for Hierarchical Matrices with Adaptive Cross Approximation [oversea]
- [1B] Organized Session "Green HPC" (--17:30) (Chair: Keiji Kimura (Waseda University))
- (K1) Approaches to the power consumption problem on the K computer
Atsuya Uno (RIKEN AICS) - (K2) The Green Supercomputer, "Shoubu" and "Suiren" MIMD Manycore Processor based, Immersion Liquid Cooled HPC System - Now and Feature -
Sunao Torii (ExaScaler) - (K3) Challenges in Power Constrained High Performance Computing
Koji Inoue, Yuichi Inadomi and Takatsugu Ono (Kyushu University)
- (K1) Approaches to the power consumption problem on the K computer
9:15∼10:45
- チュートリアル1 (Chair: Toshihiro Hanawa (The University of Tokyo))
- 「高速かつ省電力なグラフ解析とその実応用」
藤澤 克樹 (九州大学)
- 「高速かつ省電力なグラフ解析とその実応用」
11:00∼12:00
- [2A] JHPCN Organized Session "Very large-scale information systems" (Chair: Yoshiki Sato (The University of Tokyo))
- (K4) Verifying effectiveness of packet pacing on large-scale parallel programs by simulation
Hidetomo Shibamura (Institute of Systems, Information Technologies and Nanotechnologies) - (K5) A Framework for Jointing Computation Center with User-Level Management System
Hideyuki Jitsumoto (Tokyo Institute of Technology)
- (K4) Verifying effectiveness of packet pacing on large-scale parallel programs by simulation
- [2B] task scheduling (Chair: Koji Inoue (Kyushu University))
- (4) The Scalable Petascale Data-Driven Approach for the Cholesky Factorization with multiple GPUs [oversea]
Yuki Tsujita, Toshio Endo (Tokyo Institute of Technology/JST CREST) and Katsuki Fujisawa (Kyushu University/JST CREST)
出典First International Workshop on Extreme Scale Programming Models and Middleware (ESPM2 2015)
- (5) Energy-Efficient Continuous Task Scheduling for Near Real-time Periodic Tasks [oversea]
Takashi Nakada, Hiroyuki Yanagihashi (The University of Tokyo), Hiroshi Ueki, Takashi Tsuchiya, Masanori Hayashikoshi (Renesas Electronics Corporation), and Hiroshi Nakamura (The University of Tokyo)
出典IEEE International Conference on Internet of Things
- (4) The Scalable Petascale Data-Driven Approach for the Cholesky Factorization with multiple GPUs [oversea]
13:00∼14:30
- 招待講演 (Chair: Takeshi Iwashita (Hokkaido University))
- Strategies for Next Generation HPC Applications and Systems
Michael Heroux (米国サンディア国立研究所)
- Strategies for Next Generation HPC Applications and Systems
14:45∼16:15
- [3A] JHPCN Organized Session "Very large-scale numerical computation" (Chair: Kengo Nakajima (The University of Tokyo))
- (K6) High-resolution Weather Prediction Code based on High-productivity Framework for Multi-GPU computation
Takashi Shimokawabe (Tokyo Institute of Technology) - (K7) Development of numerical optimization tool for polycrystalline metallic microstructure control by using multi-phase-field method
Akinori Yamanaka (Tokyo University of Agriculture and Technology) - (K8) Interaction between a huge number of micro particles with internal degrees of freedom and turbulent mixing
Toshiyuki Gotoh (Nagoya Institute of Technology)
- (K6) High-resolution Weather Prediction Code based on High-productivity Framework for Multi-GPU computation
- [3B] computer system design (--15:45) (Chair: Takashi Nakada (The University of Tokyo))
- (6) Compact and Accurate Stochastic Circuits with Shared Random Number Sources [oversea]
Hideyuki Ichihara, Shota Ishi, Daiki Sunamori, Tsuyoshi Iwagaki and Tomoo Inoue (Hiroshima City University)
出典The 32nd IEEE International conference on computer design (ICCD)
- (7) Skywalk: a Topology for HPC Networks with Low-delay Switches [oversea]
Ikki Fujiwara, Michihiro Koibuchi (National Institute of Informatics), Hiroki Matsutani (Keio University) and Henri Casanova (University of Hawai`i at Manoa)
出典The 28th IEEE International Parallel & Distributed Processing Symposium (IPDPS 2014)
- (6) Compact and Accurate Stochastic Circuits with Shared Random Number Sources [oversea]
16:15∼16:45
16:45∼18:10
- ポスター・企業・団体展示 (Chair: Koji Inoue (Kyushu University))
- (P01) Designing a superdense ground-based camera network for airglow imaging
Shin Suzuki (Aichi University) - (P02) Flexible Virtual Computing Platform on Clouds for Scientific Workflows
Kai Liu, Kento Aida, Shigetoshi Yokoyama and Yoshinobu Masatani (National Institute of Informatics) - (P03) Comprehensive Comparison of Power-Performance Efficiency on Accelerators
Keitaro Oka, Yuichi Inadomi, Takatsugu Ono and Koji Inoue (Kyushu University) - (P04) Literate Computing for Reproducible Infrastructure
Yoshinobu Masatani, Shigetoshi Yokoyama, Nobukazu Yoshioka and Kento Aida (National Institute of informatics) - (P05) Preliminary Estimation on Automating Multi-dimensional Data Decomposition for Multi-GPU Systems
Ryotaro Sakai, Fumihiko Ino and Kenichi Hagihara (Osaka University) - (P06) Large-scale TSUNAMI simulation on New Earth Simulator
Yoshiyuki Imato, Masanobu Ishibashi (JAMSTEC), Toshitaka Baba (The University of Tokushima), Kazuto Ando, Hitoshi Uehara (JAMSTEC) and Toshihiro Kato (NEC corporation) - (P07) Filter diagonalization method for a real symmetric definite GEVP whose filter is a polynomial of a resolvent
Hiroshi Murakami (Tokyo Metropolitan University) - (P08) NSIM-ACE: A Simulator for Evaluating RDMA on Interconnection Networks
Ryutaro Susukita, Yoshiyuki Morie, Takeshi Nanri (Kyushu University) and Hidetomo Shibamura (ISIT) - (P09) Memory Control Technique of the Point‐to‐Point Communication
Takeshi Soga (ISIT) and Takeshi Nanri (Kyushu University) - (P10) Utilization and Expansion of ppOpen-AT for OpenACC
Satoshi Ohshima, Takahiro Katagiri and Masaharu Matsumoto (The University of Tokyo) - (P11) Performance Evaluation of Verified Computation for Linear Systems on Parallel Computers
Yusuke Morikura (Waseda University), Daichi Mukunoki (RIKEN), Takeshi Fukaya (Hokkaido University), Naoya Yamanaka (Teikyo Heisei University) and Shin'ichi Oishi (Waseda University) - (P12) Development of GPU-accelerated Fragment Molecular Orbital Program
Hiroaki Umeda (University of Tsukuba), Toshihiro Hanawa (The University of Tokyo), Mitsuo Shoji, Taisuke Boku and Yasuteru Shigeta (University of Tsukuba) - (P13) Consolidating memory locality information obtained from static and dynamic analysis of code for performance tuning in source code
Yukinori Sato and Toshio Endo (Tokyo Institute of Technology) - (P14) Low-latency Data Flow Management for Heterogeneous Systems
Yusuke Fujii, Shunsuke Kurumatani, Hiroyuki Takahashi and Toshio Hitaka (NTT Software Innovation Center) - (P15) Towards Parallel Graph Mining in Distributed Memory Environments using Task-Parallel Language Tascell
Shingo Okuno, Tasuku Hiraishi, Hiroshi Nakashima (Kyoto University), Masahiro Yasugi (Kyushu Institute of Technology) and Jun Sese (National Institute of Advanced Industrial Science and Technology) - (P16) Architecture of Controlling Network and Virtualized Computational Resources on SDN-enhanced Job Management System Framework
Yasuhiro Watashiba (Nara Institute of Science and Technology), Masaharu Shimizu, Susumu Date (Osaka University), Hirotake Abe (University of Tsukuba), Yoshiyuki Kido (Osaka University), Kohei Ichikawa (Nara Institute of Science and Technology), Hiroaki Yamanaka, Eiji Kawai (National Institute of Information and Communications Technology) and Shinji Shimojo (Osaka University)
- (P01) Designing a superdense ground-based camera network for airglow imaging
18:10∼20:00
9:15∼10:35
- [4A] parallelism (Chair: Osamu Tatebe (University of Tsukuba))
- ACSI2016 Outstanding Student Research Award
[27] Autotuning for Task Parallel Programs [long]
Shintaro Iwasaki and Kenjiro Taura (The University of Tokyo) - (8) An MPI-based Implementation of the Tascell Task-Parallel Programming Language [short]
Daisuke Muraoka, Masahiro Yasugi (Kyushu Institute of Technology), Tasuku Hiraishi and Seiji Umatani (Kyoto University)
- (9) Preliminary Evaluations of Probabilistic Guards for a Work-Stealing Framework [long]
Hiroshi Yoritaka (Kyushu Institute of Technology), Ken Matsui (Kyoto University), Masahiro Yasugi (Kyushu Institute of Technology), Tasuku Hiraishi and Seiji Umatani (Kyoto University)
- ACSI2016 Outstanding Student Research Award
- [4B] cloud and storage (Chair: Kenichi Kourai (Kyushu Institute of Technology))
- (10) Middleware for Building Virtual Resources on Inter-Cloud [short]
Shigetoshi Yokoyama, Yoshinobu Masatani (National Institute of Informatics), Tazro Ohta (Database Center for Life Science), Osamu Ogasawara (National Institute of Genetics), Nobukazu Yoshioka and Kento Aida (National Institute of Informatics)
- (11) Characterizing Storage Traffic of Virtual Desktop Infrastructure on Enterprise Network [long]
Chunghan Lee, Naoto Fukumoto, Hiroshi Endo, Mariko Sugawara, Tatsuma Matsuki and Tatsuo Kumano (Fujitsu Laboratories Ltd. )
- (12) RDMA-based Direct Transfer of File Data to Remote Page Cache [oversea]
Shin Sasaki, Kazushi Takahashi, Yoshihiro Oyama (The University of Electro-Communications) and Osamu Tatebe (University of Tsukuba)
出典2015 IEEE International Conference on Cluster Computing (IEEE Cluster 2015)
- (10) Middleware for Building Virtual Resources on Inter-Cloud [short]
10:50∼12:20
- [5A] performance evaluation and testing (Chair: Kenjiro Taura (The University of Tokyo))
- (13) Metrics for evaluation of parallel efficiency toward highly parallel processing [oversea]
Shigeo Orii (Tokai University)
出典Parallel Computing, 36(1), January, 2010.
- (14) A Proper Performance Evaluation System That Summarizes Code Placement Effects [oversea]
Masahiro Yasugi (Kyushu Institute of Technology), Yuki Matsuda (Kyoto University) and Tomoharu Ugawa (Kochi University of Technology)
出典The 11th ACM SIGPLAN-SIGSOFT Workshop on Program Analysis for Software Tools and Engineering (PASTE '13)
- (15) Calculation Coverage Testing in Scientific Applications [oversea]
Yoshiki Sato, Shumpei Hozumi and Shigeru Chiba (The University of Tokyo)
出典The 2015 International Symposium on Software Testing and Analysis (ISSTA 2015)
- (13) Metrics for evaluation of parallel efficiency toward highly parallel processing [oversea]
- [5B] storage and cloud (--12:10) (Chair: Yusuke Tanimura (National Institute of Advanced Industrial Science and Technology))
- (16) Design of object storage using OpenNVM for high-performance distributed file system [long]
Fuyumasa Takatsu, Kohei Hiraga and Osamu Tatebe (University of Tsukuba)
- (17) VIMM: a New Instance Management Method for Supporting OpenStack on Hitachi Server Virtualization [long]
Cheng Luo, Toshiomi Moriki and Takayuki Imada (Hitachi, Ltd.)
- (18) Efficient Write Ahead Logging with Parallel Write [short]
Komei Kamiya, Hideyuki Kawashima and Osamu Tatebe (University of Tsukuba)
- (16) Design of object storage using OpenNVM for high-performance distributed file system [long]
13:45∼15:15
- チュートリアル2 (Chair: Yoshihiro Oyama (The University of Electro-Communications))
- 「深層学習とその画像認識への応用に関する最近の研究動向」
岡谷 貴之 (東北大学)
- 「深層学習とその画像認識への応用に関する最近の研究動向」
15:25∼16:45
- [6A] communication library (Chair: Ryousei Takano (National Institute of Advanced Industrial Science and Technology))
- (19) Evaluation of On-Demand Message-Passing Module over RDMA Network [short]
Takeshi Nanri (Kyushu University) and Takeshi Soga (ISIT Kyushu)
- ACSI2016 Outstanding Research Award
[24] A proposal of GMPI: GPU self MPI for GPU clusters [long]
Yuta Kuwahara (University of Tsukuba), Toshihiro Hanawa (The University of Tokyo) and Taisuke Boku (University of Tsukuba) - (20) ACPdl: Data-Structure and Global Memory Allocator Library over a Thin PGAS-Layer [oversea]
Yuichiro Ajima, Takafumi Nose, Kazushige Saga, Naoyuki Shida and Shinji Sumimoto (Fujitsu Limited)
出典First International Workshop on Extreme Scale Programming Models and Middleware (ESPM2 2015)
- (19) Evaluation of On-Demand Message-Passing Module over RDMA Network [short]
- [6B] graph analysis (--16:25) (Chair: Daisuke Takahashi (University of Tsukuba))
- (21) Efficient Use of Hardware Transactional Memory for Parallel Mesh Generation [oversea]
Tetsu Kobayashi, Shigeyuki Sato and Hideya Iwasaki (The University of Electro-Communications)
出典The 44th International Conference on Parallel Processing (ICPP 2015)
- (22) Interactive high-dimensional visualization of social graphs [oversea]
Ken Wakita (Tokyo Institute of Technology/ JST CREST), Masanori Takami (Tokyo Institute of Technology) and Hiroshi Hosobe (Hosei University)
出典The 8th IEEE Pacific Visualization Symposium (PacificVis 2015)
- (21) Efficient Use of Hardware Transactional Memory for Parallel Mesh Generation [oversea]
16:45∼17:00