HANAMI High-Level Symposium | 3rd Edition

Mohamed Wahib is a team principal (PI) of the “High Performance Artificial Intelligence Systems Research Team” at RIKEN Center for Computational Science (R-CCS), Kobe, Japan. Prior to that he worked as a senior scientist at AIST/TokyoTech Open Innovation Laboratory, Tokyo, Japan. His research interests revolve around the central topic of high-performance programming systems, in the context of HPC and AI. He is actively working on several projects including AI-based science, as well as high-level frameworks for programming traditional scientific applications.

Talk: ML Surrogates in Scientific Applications

Dr. Shin-ichiro Shima has been at the University of Hyogo since 2011 and became a professor in 2024. After completing his Ph.D. in nonlinear dynamics at Kyoto University in 2005, he became a research scientist at the Earth Simulator Center, Japan Agency for Marine-Earth Science and Technology. With an eye on the future of supercomputers, he developed novel numerical algorithms for multiscale-multiphysics phenomena at this supercomputer center and constructed the super-droplet method (SDM), a Lagrangian particle-based algorithm for cloud microphysics. Since then, he has been working on this topic to explore the potential of the particle-based cloud modeling method.

Talk: Multiscale Cloud Microphysics Modeling from Aerosols to Convection with the Super-Droplet Method

Claudia Filippi is a Professor in Computational Chemical Physics at the University of Twente, the Netherlands. She received her PhD from Cornell University in 1996. After a postdoc at the University of Illinois at Urbana-Champaign, she held academic positions at University College Cork and Leiden University before joining the University of Twente in 2009. She has made significant contributions to the development of quantum Monte Carlo methods, connecting accurate electronic-structure theory with multiscale modelling and applications to molecular and materials systems. She coordinated the European TREX Center of Excellence in Exascale Computing, leading the development of efficient, scalable software solutions for stochastic quantum chemistry.

Talk: Quantum Monte Carlo forces: from reference data to machine-learned models

Clément Richefort is a Postdoctoral Researcher at the Jülich Supercomputing Centre (JSC), where he specialises in bridging theoretical numerical linear algebra with HPC. He holds a PhD from the Université de Bordeaux, and his expertise spans sparse multigrid methods, iterative solvers, and GPU programming, developed through research collaborations with CEA and Lawrence Livermore National Laboratory. Currently, he works within the EU-Japan HANAMI collaboration, optimising the multi-GPU ChASE iterative eigensolver on the JUPITER exascale system for materials science applications. His latest research focuses on integrating the Ozaki Scheme II into ChASE to enable high-accuracy FP-GEMM operations utilising INT8 Tensor Cores.

Talk: Leveraging Low-Precision AI Hardware for FP64 Eigensolvers : Integrating the Ozaki-scheme II into ChASE

Katsuhisa Ozaki is a Professor in the Department of Mathematical Sciences at Shibaura Institute of Technology. He received his Ph.D. in Engineering from Waseda University in 2007. He was an Assistant Professor at Waseda University from April 2007 to March 2008, and a full-time Visiting Lecturer from April 2008 to March 2010. At Shibaura Institute of Technology, he served as an Assistant Professor from April 2010 to March 2013 and as an Associate Professor from April 2013 to March 2019, and has been a Professor since April 2019. His research interests include reliable computing, with a particular focus on rounding error analysis in finite-precision arithmetic. His main research area is numerical linear algebra, where he develops fast and accurate algorithms for reliable scientific computing.

Talk: Accuracy-Tunable Matrix Computation Using Low-Precision Arithmetic: Ozaki Schemes and Applications