Directions
Facebook
Twitter
Instagram
YouTube
Japanese
Contact Us
TOP
Research
Research Teams
High Performance Artificial Intelligence Systems Research Team
High Performance Artificial Intelligence Systems Research Team
Japanese
Team Leader Satoshi Matsuoka
matsu[at]acm.org (Lab location: Tokyo)- Please change [at] to @
- 2019
- Team Leader, High Performance Artificial Intelligence Systems Research Team, R-CCS, RIKEN (-present)
- 2018
- Director, R-CCS, RIKEN (-present)
- Specially Appointed Professor, Tokyo Tech (-present)
- 2017
- Director, Real World Big Data Computing Open Innovation Laboratory (RWBC-OIL), AIST and Tokyo Tech
- 2000
- Full Professor, Global Scientific Information and Computing Center (GSIC), the Tokyo Institute of Technology
- 1993
- Ph. D. from the University of Tokyo
Keyword
- High Performance Artificial Intelligence Systems
- Scalable Deep Learning
- Performance Modeling of AI Systems e.g. Deep Learning
- Acceleration of Advanced Deep Learning Algorithms
- Convergence of AI and Simulation
Research summary
The High Performance Artificial Intelligence Systems Research Team is an R-CCS laboratory focusing on convergence of HPC and AI, namely high performance systems, software, and algorithms research for artificial intelligence/machine learning. In collaboration with other research institutes in HPC and AI-related research in Japan as well as globally, it seeks to develop next-generation AI technology that will utilize state-of-the-art high-performance computation facilities, including the post-K. Specifically, we conduct research on next-generation AI systems by focusing on the following topics:
- Extreme speedup and scalability of deep learning: Achieve extreme scalability of deep learning in large-scale supercomputing environments including the post-K, extending the latest algorithms and frameworks for deep learning.
- Performance analysis of deep learning: Accelerate computational kernels for AI over the state-of-the-art hardware architectures by analyzing algorithms for deep learning and other machine learning/AI, measuring their performance and constructing their performance models.
- Acceleration of modern AI algorithms: Accelerate advanced AI algorithms, such as ultra-deep neural networks and high-resolution GAN over images, those that require massive computational resources, using extreme-scale deep learning systems.
- Acceleration of HPC algorithms using machine learning: Accelerate HPC algorithms and applications using empirical models based on machine learning.
Representative papers
- Kazuki Osawa, Yohei Tsuji, Yuichiro Ueno, Akira Naruse, Rio Yokota, and Satoshi Matsuoka.:
"Large-scale Distributed Second-order Optimization Using Kronecker-factored Approximate Curvature for Deep Convolutional Neural Networks"
2019 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2019), to appear. - Yusuke Nagasaka, Akira Nukada, Ryosuke Kojima, and Satoshi Matsuoka.:
"Batched Sparse Matrix Multiplication for Accelerating Graph Convolutional Networks"
The 19th Annual IEEE/ACM International Symposium in Cluster, Cloud, and Grid Computing (CCGrid 2019), to appear. - Shweta Salaria, Aleksandr Drozd, Artur Podobas, and Satoshi Matsuoka.:
"Learning Neural Representations for Predicting GPU Performance"
ISC High Performance 2019 (ISC'19), to appear. - Pak Markthub, and Mehmet E. Belviranli, Seyong Lee, Jeffrey Vetter, and Satoshi Matsuoka.:
"DRAGON: Breaking GPU Memory Capacity Limits with Direct NVM Access"
Proceedings of the ACM/IEEE International Conference for High Performance Computing, Networking, Storage, and Analysis (SC'18), pp 32:1--32:13. - Yosuke Oyama, Tal Ben-Nun, Torsten Hoefler, and Satoshi Matsuoka.:
"Accelerating Deep Learning Frameworks with Micro-batches"
2018 IEEE International Conference on Cluster Computing (CLUSTER 2018) pp 402-412. - Yosuke Oyama, Akihiro Nomura, Ikuro Sato, Hiroki Nishimura, Yukimasa Tamatsu, and Satoshi Matsuoka.:
"Predicting Statistics of Asynchronous SGD Parameters for a Large-Scale Distributed Deep Learning System on GPU Supercomputers"
2016 IEEE International Conference on Big Data (Big Data 2016), pp. 66-75.
