Hiroshi Inoue           

井上 拓 (Hiroshi Inoue)

• 2021年11月: IEEE BigData 2021に，SIMDを使うキャシュリプレースメントアルゴリズムについての論文が採択されました．(short paper)
• 2021年5月: ISCA 2021に，DNNアクセラレータについての論文が採択されました（共著）
• 2020年11月: Proceedings of the IEEEに，RAPID AI アクセラレータについての記事が掲載されました（共著）
• 2019年8月: IEEE Microに，DNNアクセラレータ向けのソフトウェアスタックについての記事が掲載されました（共著）
• 2019年5月: Deep learning向けのDropoutの拡張についての論文を arXiv.org で公開しました．
• 2019年3月: COOL Chips XXIIに，DNNアクセラレータ向けのコンパイラ構成についての論文が採択されました（共著）[ paper ]
• 2018年12月: AISTATS 2019に，DeepLearning向けのensemble技術についての論文が採択されました．[ paper at publisher web page ]
• 2018年1月: Deep learning向けの新しいdata augmentation技術についての論文を arXiv.org で公開しました．
• 2017年10月: IEEE BigData 2017に計算量の削減による3次元CT再構成の高速化についての論文が採択されました．[ paper, slides ]
• 2017年1月: IPDPS 2017にApache Sparkでのコード最適化に関する論文が採択されました（共著）
• 2015年5月: VLDB 2015 (PVLDB Vol.8, No 11)にSIMD命令を用いる構造体向けのソートアルゴリズムについての論文が採択されました． [ paper at publisher web page, slides]
• 2014年9月: VLDB 2015 (PVLDB Vol.8, No 3)にSIMD命令を用いるset intersectionの高速化についての論文が採択されました．[ paper at publisher web page, slides]

コンパイラ，言語ランタイム，並列アルゴリズムなど
現在はハードウェアの機能を活用して，様々なワークロードを高速化する手法を中心に研究を行っています．主な対象としてはマルチコアプロセッサのスレッド並列度の活用や，PowerPCやCellプロセッサのSIMD命令，ハードウェアパフォーマンスカウンタの活用など．

2016年1月 Outstanding Research Award at ACSI 2016
2015年1月 Outstanding Research Award at ACSI 2015
2008年7月 情報処理学会 山下記念研究賞
2000年3月 慶應義塾大学 義塾賞

Conference Papers

• Hiroshi Inoue, "Multi-step LRU: SIMD-based Cache Replacement for Lower Overhead and Higher Precision", IEEE International Conference on Big Data (IEEE BigData 2021), online, December 15-18, 2021. (short paper, Acceptance rate: (97+96)/486 = 39.7%)
• Swagath Venkataramani, Vijayalakshmi Srinivasan, Wei Wang, Sanchari Sen, Jintao Zhang, Ankur Agrawal, Monodeep Kar, Shubham Jain, Alberto Mannari, Hoang Tran, Yulong Li, Eri Ogawa, Kazuaki Ishizaki, Hiroshi Inoue, Marcel Schaal, Mauricio Serrano, Jungwook Choi, Xiao Sun, Naigang Wang, Chia-Yu Chen, Allison Allain, James Bonano, Nianzheng Cao, Robert Casatuta, Matthew Cohen, Bruce Fleischer, Michael Guillorn, Howard Haynie, Jinwook Jung, Mingu Kang, Kyu-hyoun Kim, Siyu Koswatta, Saekyu Lee, Martin Lutz, Silvia Mueller, Jinwook Oh, Ashish Ranjan, Zhibin Ren, Scot Rider, Kerstin Schelm, Michael Scheuermann, Joel Silberman, Jie Yang, Vidhi Zalani, Xin Zhang, Ching Zhou, Matt Ziegler, Vinay Shah, Moriyoshi Ohara, Pong-Fei Lu, Brian Curran, Sunil Shukla, Leland Chang, Kailash Gopalakrishnan, "RaPiD: AI Accelerator for Ultra-Low Precision Training and Inference", International Symposium on Computer Architecture (ISCA 2021), June 14-16, 2021.
• Eri Ogawa, Kazuaki Ishizaki, Hiroshi Inoue, Swagath Venkataramani, Jungwook Choi, Wei Wang, Vijayalakshmi Srinivasan, Moriyoshi Oharaand Kailash Gopalakrishn, "A Compiler for Deep Neural Network Accelerators to Generate Optimized Code for a Wide Range of Data Parameters from a Hand-crafted Computation Kernel", IEEE Symposium on Low-Power and High-Speed Chips (COOL Chips XXII), Yokohama, Japan, April 17-29, 2019.
• Hiroshi Inoue, "Adaptive Ensemble Prediction for Deep Neural Networks based on Confidence Level", The 22nd International Conference on Artificial Intelligence and Statistics (AISTATS 2019), Naha, Okinawa, Japan, April 16-18, 2019. (Acceptance rate: 360/1111 = 32.4%)
• Hiroshi Inoue, "Fast Interpolation of Grid Data at a Non-Grid Point", IEEE International Conference on Big Data (IEEE BigData 2017), Boston, MA, USA, December 11-14, 2017. (Acceptance rate: 79/437 = 18.1%)
• Jan Wróblewski, Kazuaki Ishizaki, Hiroshi Inoue and Moriyoshi Ohara, "Accelerating Spark Datasets by inlining deserialization", 31st IEEE International Parallel & Distributed Processing Symposium (IPDPS 2017), May 29-June 2, 2017. (Acceptance rate: 116/508 = 22.8%)
• Masaru Ito, Hiroshi Inoue and Kenjiro Taura, "Fragmented BWT: Extended BWT for full-text indexing", International Symposium on String Processing and Information Retrieval (SPIRE 2016), Beppu, Japan, October 18-20, 2016. (Acceptance rate: 25/46 = 54.3%)
• Hiroshi Inoue, "How SIMD Width Affects Energy Efficiency: A Case Study on Sorting", IEEE Symposium on Low-Power and High-Speed Chips (COOL Chips XIX), Yokohama, Japan, April 20-22, 2016. (Acceptance rate: 11/22 = 50.0%)
• Hiroshi Inoue, "Efficient Tomographic Reconstruction For Commodity Processors with Limited Memory Bandwidth", The 2016 IEEE International Symposium on Biomedical Imaging (ISBI 2016), Prague, Czech Republic, April 13-16, 2016. pp 747-750. (Acceptance rate: 337/649 = 51.9%) photos
• Hiroshi Inoue and Kenjiro Taura, "SIMD- and Cache-Friendly Algorithm for Sorting an Array of Structures", PVLDB 8(11), pp 1274-1285, presented in 41st International Conference on Very Large Data Bases (VLDB 2015). Kohala Coast, Hawaii, USA, August 31 - September 4, 2015.  (Acceptance rate: 151/710 = 21.3%) photos
• Hiroshi Inoue, Moriyoshi Ohara and Kenjiro Taura, "Faster Set Intersection with SIMD instructions by Reducing Branch Mispredictions", PVLDB 8(3), pp 293-304, presented in 41st International Conference on Very Large Data Bases (VLDB 2015). Kohala Coast, Hawaii, USA, August 31 - September 4, 2015.  (Acceptance rate: 151/710 = 21.3%) photos
• Hiroshi Inoue and Toshio Nakatani, "Adaptive SMT Control for More Responsive Web Applications", 2014 IEEE International Symposium on Workload Characterization (IISWC 2014). Raleigh, North Carolina, USA. pp 41-50. October 26-28, 2014. (Acceptance rate: 22/80 = 27.5%) photos
  
• Takuya Nakaike, Hiroshi Inoue, Toshio Suganuma, and Moriyoshi Ohara, "Characterization of Call-Graph Profiles in Java Workloads", 2014 IEEE International Symposium on Workload Characterization (IISWC 2014). Raleigh, North Carolina, USA. pp 161-170. October 26-28, 2014. (Acceptance rate: 22/80 = 27.5%)
• Hiroshi Inoue, Hiroshige Hayashizaki, Peng Wu, and Toshio Nakatani, "Adaptive Multi-Level Compilation in a Trace-based Java JIT Compiler", 2012 ACM Object-Oriented Programming, Systems, Languages & Applications (SPLASH/OOPSLA 2012). Tucson, Arizona, USA. pp 179-194, October 19-26, 2012. (Acceptance rate: 59/228 = 25.9%) photos
• Hiroshi Inoue and Toshio Nakatani, "Identifying the Sources of Cache Misses in Java Programs Without Relying on Hardware Counters", 2012 International Symposium on Memory Management (ISMM 2012). Beijing, China. pp 133-142. June 15-16, 2012. (Acceptance rate: 12/30 = 40.0%) photos
• Peng Wu, Hiroshige Hayashizaki, Hiroshi Inoue, and Toshio Nakatani, "Reducing Trace Selection Footprint for Large-scale Java Applications with no Performance Loss", 2011 ACM Object-Oriented Programming, Systems, Languages & Applications (SPLASH/OOPSLA 2011). Portland, Oregon, USA. pp 789-804. October 22-27, 2011. (Acceptance rate: 61/166 = 36.7%)
• Hiroshi Inoue, Hiroshige Hayashizaki, Peng Wu, and Toshio Nakatani, "A Trace-based Java JIT Compiler Retrofitted from a Method-based Compiler", 2011 9th Annual IEEE/ACM International Symposium on Code Generation and Optimization (CGO 2011). Chamonix, France. pp. 246-256, April 2-6, 2011. (Acceptance rate: 28/105 = 26.7%) photos
• Hiroshige Hayashizaki, Peng Wu, Hiroshi Inoue, Mauricio Serrano and Toshio Nakatani, "Improving the Performance of Trace-based Systems by False Loop Filtering", Sixteenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS 2011). Newport Beach, California, USA. pp. 405-418, March 5-11, 2011. (Acceptance rate: 32/152 = 21.1%)
• Hiroshi Inoue and Toshio Nakatani, "Performance of Multi-Process and Multi-Thread Processing on Multi-core SMT Processors", 2010 IEEE International Symposium on Workload Characterization (IISWC 2010). Atlanta, Georgia, USA. pp 209-218. December 2-4, 2010. (Acceptance rate: 21/56 = 37.5%) photos
• Hiroshi Inoue and Toshio Nakatani, "How a Java VM Can Get More from a Hardware Performance Monitor", ACM SIGPLAN 2009 International Conference on Object-Oriented Programming, Systems, Languages, and Applications (OOPSLA 2009). Orlando, Florida, USA. pp. 137-154. October 25-29, 2009. (Acceptance rate: 25/144 = 17.4%) photos
• Hiroshi Inoue, Hideaki Komatsu, and Toshio Nakatani, "A Study of Memory Management for Web-based Applications on Multicore Processors", ACM SIGPLAN 2009 Conference on Programming Language Design and Implementation (PLDI 2009). Dublin, Ireland. pp. 386-396. June 15-20, 2009. (Acceptance rate: 41/196 = 20.9%) photos
• Hiroshi Inoue, Takao Moriyama, Hideaki Komatsu, and Toshio Nakatani, "AA-Sort: A New Parallel Sorting Algorithm for Multi-Core SIMD Processors", IEEE The Sixteenth International Conference on Parallel Architectures and Compilation Techniques (PACT 2007). Brasov, Romania. pp. 189-198. Sept. 15-19, 2007 (Acceptance rate: 34/175 = 19.4%)
• Jessica H. Tseng, Hao Yu, Shailabh Nagar, Niteesh Dubey, Hubertus Franke, Pratap Pattnaik, Hiroshi Inoue, and Toshio Nakatani, "Performance Studies of Commercial Workloads on a Multi-core System", 2007 IEEE International Symposium on Workload Characterization (IISWC 2007), 2007.
• Moriyoshi Ohara, Hangu Yeo, Frank Savino, Giridharan Iyengar, Leiguang Gong, Hiroshi Inoue, Hideaki Komatsu, Vadim Sheinin, and Shahrokh Daijavad, "Accelerating Mutual-Information-Based Linear Registration on the Cell Broadband Engine Processor", 2007 IEEE International Conference on Multimedia and Expo (ICME 2007), 2007.
• Moriyoshi Ohara, Hangu Yeo, Frank Savino, Giridharan Iyengar, Leiguang Gong, Hiroshi Inoue, Hideaki Komatsu, Vadim Sheinin, Shahrokh Daijavad, and Bradley Erickson, "Real-time Mutual-information-based Linear Registration on the Cell Broadband Engine Processor", Fourth IEEE Symposium on Biomedical Imaging (ISBI 2007), 2007.
• Motohiro Kawahito, Hideaki Komatsu, Takao Moriyama, Hiroshi Inoue, and Toshio Nakatani, "A New Idiom Recognition Framework for Exploiting Hardware-Assist Instructions", Twelfth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS XII), 2006.
• 井上 拓, 森山 孝男, 小松 秀昭, 中谷 登志男, "VMX命令セットを用いる高速なソートアルゴリズム", 情報処理学会ハイパフォーマンスコンピューティングと計算科学シンポジウム, 2006.
• 井上 拓, 森山 孝男, 根岸 康, 小原 盛幹, "SMT技術を用いるCPU上で動作する計算資源予約システム", 情報処理学会コンピュータシステムシンポジウム, 2003.
• Hiroshi Inoue, Yohei Sato, and Koichi Hishida, "Directional Scale Dependency on Force Coupling for Dispersed Two-Phase Turbulent Flows", Second International Symposium on Turbulence and Shear Flow Phenomena (TSFP2), 2001.

Journal Papers

• Motohiro Kawahito, Hideaki Komatsu, Takao Moriyama, Hiroshi Inoue, and Toshio Nakatani, "Idiom Recognition Framework using Topological Embedding", ACM Transactions on Architecture and Code Optimization (TACO), Vol. 10(3), 2013.
• Hiroshi Inoue, Takao Moriyama, Hideaki Komatsu, and Toshio Nakatani, "A high-performance sorting algorithm for multicore single-instruction multiple-data processors", Software: Practice and Experience, Vol. 42(6), pp. 753-777, 2012.
• 井上 拓, 小松 秀昭, 中谷 登志男, "SIMD命令を用いるUTF-8文字列デコード処理の高速化", 情報処理学会論文誌:プログラミング, Vol.1 (2), pp. 1-8, 2008.
• Moriyoshi Ohara, Hiroshi Inoue, Yukihiko Sohda, Hideaki Komatsu, and Toshio Nakatani, "MPI microtask for programming the Cell Broadband Engine processor", IBM Systems Journal, Vol 45 (1), pp. 85-, 2006.
• 井上 拓, 森山 孝男, 小松 秀昭, 中谷 登志男, "VMX命令セットを用いる高速なソートアルゴリズム", 情報処理学会論文誌:ACS, Vol 47 (ACS 14), pp. 105-, 2006.
• 井上 拓, 森山 孝男, 根岸 康, 小原 盛幹, "SMT技術を用いるCPU上で動作する計算資源予約システム", 情報処理学会論文誌:ACS, Vol.45 (ACS 5), pp. 21-, 2004.

招待講演

• 井上 拓, "RaPiD AI アクセラレータ向けソフトウェアスタックのデザインと実装", IEICE-CPSY/DC, IPSJ-ARC Hot SPring Annual meeting 2021 (HotSPA2021). 2021.
• Hiroshi Inoue, "Efficient Optimization of Diameter and Average Shortest Path Length of a Graph using Path Count Index", Graph Golf Workshop at The Fourth International Symposium on Computing and Networking (CANDAR'16). Hiroshima, Japan, 2016.
• Hiroshi Inoue, Hiroshige Hayashizaki, Peng Wu, and Toshio Nakatani, "Adaptive Multi-Level Compilation in a Trace-based Java JIT Compiler", 日本ソフトウェア科学会第30回大会, September 10-13, 2013.
• Hiroshi Inoue, "A Trace-based Java JIT Compiler for Large-scale Applications", The 6th workshop on Virtual Machines and Intermediate Languages (VMIL2012). Tucson, Arizona, USA. 2012.

Workshop papers

• Peng Wu, Hiroshige Hayashizaki, Hiroshi Inoue, and Toshio Nakatani, "Reducing Trace Selection Footprint for Large-scale Java Applications without Performance Loss", 10th Workshop on Compiler-Driven Performance, 2011.
• Peng Wu, Hiroshige Hayashizaki, and Hiroshi Inoue, "Understand the Building Blocks of Trace Selection for a Trace-driven Language Compiler", 9th Workshop on Compiler-Driven Performance, 2010.
• 井上 拓, 小松 秀昭, 中谷 登志男, "SIMD命令を用いるUTF-8文字列デコード処理の高速化", 情報処理学会 第68回プログラミング研究会, 2008. (山下記念研究賞受賞)
• Moriyoshi Ohara, Hangu Yeo, Frank Savino, Giridharan Iyengar, Leiguang Gong, Hiroshi Inoue, Hideaki Komatsu, Vadim Sheinin, and Shahrokh Daijavad, "Accelerating medical image registration on the Cell broadband engine processor", Second Workshop on Real Time and Interactive Digital Media Supercomputing (RIDMS-2), 2007.
• Daniel Citron, Hiroshi Inoue, Takao Moriyama, Motohiro Kawahito, Hideaki Komatsu, and Toshio Nakatani, "Exploiting the AltiVec Unit for Commercial Applications", Workshop on Computer Architecture Evaluation using Commercial Workloads, 2006.
• 井上 拓, "SMT環境下におけるCPU資源予約", ソフトウェア科学会プログラミングおよびシステムに関するワークショップ (SPA), 2003.

Other publications

• Hiroshi Inoue, "Multi-step LRU: SIMD-based Cache Replacement for Lower Overhead and Higher Precision", arXiv:2112.09981 [cs.NI], 2021
• Swagath Venkataramani, Xiao Sun, Naigang Wang, Chia-yu Chen, Jungwook Choi, Mingu Kang, Ankur Agarwal, Jinwook Oh, Shubham Jain, Tina Babinsky, Nianzheng Cao, Thomas Fox, Bruce Fleischer, George Gristede, Michael Guillorn, Howard Haynie, Hiroshi Inoue, Kazuaki Ishizaki, Michael Klaiber, Shih-hsien Lo, Gary Maier, Silvia Mueller, Michael Scheuermann, Eri Ogawa, Marcel Schaal, Mauricio Serrano, Joel Silberman, Christos Vezyrtzis, Wei Wang, Fanchieh Yee, Jintao Zhang, Matthew Ziegler, Ching Zhou, Moriyoshi Ohara, Pong-fei Lu, Life Brian Curran, Sunil Shukla, Vijayalakshmi Srinivasan, Leland Chang, And Kailash Gopalakrishnan, "Efficient AI System Design With Cross-Layer Approximate Computing", Proceedings of the IEEE, 2020
• Swagath Venkataramani, Jungwook Choi, Vijayalakshmi Srinivasan, Wei Wang, Jintao Zhang, Marcel Schaal, Mauricio Serrano, Kazuaki Ishizaki, Hiroshi Inoue, Eri Ogawa, Moriyoshi Ohara, Leland Chang, Kailash Gopalakrishnan, "DeepTools: Compiler and Execution Runtime Extensions for RaPiD AI Accelerator", IEEE Micro, 2019
• Hiroshi Inoue, "Multi-Sample Dropout for Accelerated Training and Better Generalization",  arXiv:1905.09788 [cs.NE], 2019
• Hiroshi Inoue, "Data Augmentation by Pairing Samples for Images Classification",  arXiv:1801.02929 [cs.LG], 2018
• Masaru Ito, Hiroshi Inoue, Megumi Ito and Moriyoshi Ohara, "FBWTMEM : computing maximal exact matches with FBWT", IBM Research Report RT0981, 2017.
• Hiroshi Inoue, "Bring Apache Spark Closer to Accelerators", Workshop on Recent Topics in High Performance Computing (HPCAsia PC Workshop), 2017
• Hiroshi Inoue, "Fast and Accurate Inference with Adaptive Ensemble Prediction in Image Classification with Deep Neural Networks", IBM Research Report RT0978, also arXiv:1702.08259 [cs.LG], 2017
• Blog post at Spark.tc, Gita Koblents, Kazuaki Ishizaki, Hiroshi Inoue, "Bringing Apache Spark Closer to SIMD and GPU", http://www.spark.tc/simd-and-gpu/, 2016
• 仲池 卓也, 井上 拓, "FCRC 2015/ISCA 2015参加報告", コンピュータ ソフトウェア Vol. 33, No. 1, pp. 22-26, 2016
• IBM Whitepaper, "Real-Time Mutual-Information-Based Linear Registration on the Cell Broadband Engine Processor", 2007.
• Hiroshi Inoue, Takao Moriyama, Yasushi Negishi, and Moriyoshi Ohara, "CPU Resource Reservation for Simultaneous Multi-Thread Systems", IBM Research Report, 2006.

プログラミングコンテスト

• ACM ICFP programming contest 2013, 8th place in the main division and 5th place in the lightning division.

• CGO 2023 PC member
• IEEE Big Data 2022 PC member
• IEEE ICMLA 2022 PC member
• IEEE Big Data 2021 PC member
• IEEE ICMLA 2021 PC member
• IEEE Big Data 2020 PC member
• IJCAI 2020 PC member
• IEEE Big Data 2019 PC member
• IJCAI 2019 PC member
• HPCAsia 2019 PC member
• CGO 2019 PC member
• IEEE Big Data 2018 PC member
• IEEE Cluster 2018 PC member (Programming and System Software area)
• HPCAsia 2018 PC member
• IEEE Big Data 2017 PC member
• IEEE Cluster 2017 PC member (Programming and System Software area)
• ICS 2016 PC member photos
• SAC 2016 Programming Languages track PC member
• IISWC 2015 PC member photos
• SAC 2015 Programming Languages track PC member
• PLDI 2015 External review committee member
• SAC 2014 Programming Languages track PC member
• SC13 PC member (programming systems track) photos
• PLDI 2013 External review committee member
• ASPLOS 2013 External review committee member
• Reviewer for ACM TACO, ACM TECS, ACM TOMS, ACM JETC, IEEE TC, IEEE TPDS, VLDB Journal, Software: Practice and Experience, 情報処理学会論文誌 etc.

Society Governance

• 日本ソフトウェア科学会 理事 (2019年6月-)

2000年3月 慶應義塾大学 理工学部 システムデザイン工学科卒業
2002年3月 慶應義塾大学大学院 理工学研究科 総合デザイン工学専攻 修士課程修了
　(修士論文：分散系二相乱流中の粒子による乱流構造変化 指導教官：菱田公一教授)
2016年3月 東京大学 情報理工学系研究科 電子情報学専攻 博士課程修了 博士（情報理工学）
　(博士論文：SIMD命令の効率的な活用による非数値計算アプリケーションの高速化 指導教官：田浦健次朗教授)
2002年4月〜現在 日本IBM(株)入社 現在まで同社東京基礎研究所に在籍