Back

Minisymposium Presentation

What and How Would we Build the Future Eigenvalue Solver?

Wednesday, June 18, 2025
15:30
-
16:00
CEST
Climate, Weather and Earth Sciences
Climate, Weather and Earth Sciences
Climate, Weather and Earth Sciences
Chemistry and Materials
Chemistry and Materials
Chemistry and Materials
Computer Science and Applied Mathematics
Computer Science and Applied Mathematics
Computer Science and Applied Mathematics
Humanities and Social Sciences
Humanities and Social Sciences
Humanities and Social Sciences
Engineering
Engineering
Engineering
Life Sciences
Life Sciences
Life Sciences
Physics
Physics
Physics

Presenter

Toshiyuki
Imamura
-
RIKEN

Toshiyuki Imamura is a team leader of Large-scale Parallel Numerical Computing Technology Team at RIKEN R-CCS and is responsible for developing numerical libraries on Fugaku. He received his Diploma and Doctorate in Applied Systems and Sciences from Kyoto University in 1993 and 2000. He was a Researcher at CCSE, JAERI (1996-2003), a visiting scientist at HLRS (2002), and an associate professor at the University of Electro-Communications (2003-2012). His research team has developed and maintained high-performance numerical software, especially for numerical linear algebra, on the K computer and the Fugaku supercomputer, and these activities will continue on the Fugaku next system. His research interests include HPC, performance autotuning technology, parallel eigenvalue computation, and developing the high-performance parallel eigenvalue library EigenExa and its variations towards emerging GPU systems. His research group won the HPL-MpX ranking with the first exascale benchmark record 2.0EFLOPS using the full Fugaku system in 2020-2021, and he was nominated as the Gordon Bell Prize finalist member in SC05, SC06, and SC20 with significant contribution of large-scale parallel eigensolver for scientific application codes.

Description

In the Japanese computational science community, which has developed the K computer and Fugaku, the high demand for large-scale eigenvalue calculations in condensed material science has prompted updates to numerical software. Capability Computing is a crucial method for effectively addressing challenging large-scale systems and is a standard scientific tool. On the other hand, an evolving approach is also essential from the perspective of Capacity Computing, which manages large batches of eigenvalue computations. We aim to develop statistical, ensemble, and AI-enabled computational frameworks that leverage advanced approximation algorithms, cutting-edge system runtimes, and software frameworks such as Kokkos, IRIS, C++, Python, and Julia. Our goal is to discuss the development of the solver and present a roadmap that connects the creation of next-generation mathematical software with this framework and next-generation computers, encouraging participants to engage in conversation.

Authors