Supercomputing Division, Information Technology Center, The University Tokyo

45th ASE Seminar (Advanced Supercomputing Environment): International Workshop on "Integration of Simulation/Data/Learning and Beyond" (Hybrid)

International Workshop on "Integration of Simulation/Data/Learning and Beyond" will be co-hosted by JSPS Grant-in-Aid for Scientific Research (S): Innovative Methods for Scientific Computing in the Exascale Era by Integrations of (Simulation+Data+Learning) (19H05662, Leading PI: Kengo Nakajima (University of Tokyo)), Information Technology Center/the University of Tokyo, and Joint Usage/Research Center for Interdisciplinary Large-scale Information Infrastructures (JHPCN), as the 45th ASE Seminar (Advanced Supercomputing Environment), Information Technology Center, the University of Tokyo.
Recently, supercomputing has been changing dramatically. Integration/convergence of Simulation/Data/Learning (S+D+L) is important towards Society 5.0 proposed by Japanese Government, which enables integration of cyber space & physical space. We propose an innovative method for computational science for sustainable promotion of scientific discovery by supercomputers in the Exascale Era by combining (Simulation + Data + Learning (S+D+L)) in the project, and have developed a software platform "h3-Open-BDEC" for integration of (S+D+L) and evaluate the effects of integration of (S+D+L) on the Wisteria/BDEC-01 system. The h3-Open-BDEC is designed for extracting the maximum performance of the supercomputers with minimum energy consumption focusing on (1) innovative method for numerical analysis with high-performance/high-reliability/power-saving based on the new principle of computing by adaptive precision, accuracy verification and automatic tuning, (2) Hierarchical Data Driven Approach (hDDA) based on machine learning, and (3) Software for heterogeneous systems, such as Wisteria/BDEC-01. Integration of (S+D+L) by h3-Open-BDEC enables significant reduction of computations and power consumption, compared to those by conventional simulations. This year is the final year of the project, and at this workshop we will introduce the research results to date as well as discuss future prospects. We will also invite Professor Richard Vuduc (Georgia Institute of Technology, USA) and Professor Xing Cai (Simula Research Laboratory, Norway) to give keynote talks, and participate in discussions. Please join us. Participation is free, but please be sure to register in advance. Please note that on-site participation will be limited to the first 35 people, so please apply early. All presentations will be in English.

45th ASE Seminar

Date & Time: November 29, 2023 (Wednesday), 13:00 - 17:45
Style: Hybrid of Online (Zoom) and On-Site (T412, Information Technology Center, Kashiwa II Campus, the University of Tokyo)
   Kashiwa Campus Map
   Accessibility Map for Kashiwa Campus
Co-Hosted by: JSPS Grant-in-Aid for Scientific Research (S): Innovative Methods for Scientific Computing in the Exascale Era by Integrations of (Simulation+Data+Learning) (19H05662, Leading PI: Kengo Nakajima (University of Tokyo)), Information Technology Center/The University of Tokyo, Joint Usage/Research Center for Interdisciplinary Large-scale Information Infrastructures (JHPCN)

Program

45th ASE Seminar (Advanced Supercomputing Environment): International Workshop on "Integration of Simulation/Data/Learning and Beyond"
13:00-17:45, November 29th (Wednesday), 2023

13:00 - 13:30 Kengo Nakajima (The University of Tokyo)
Integration of Simulation/Data/Learning and Beyond
Presentation Material
13:30 - 13:45 Takeshi Iwashita (Online) (Hokkaido University, Kyoto University)
Research activities of Hokkaido University group for next-generation linear solvers -mixed precision computing, accelerators, subspace correction techniques-
Presentation Material
13:45 - 14:00 Takahiro Katagiri (Nagoya University)
Exploring AI for Auto-tuning through Sparse Matrix Image Information
Presentation Material
14:00 - 14:15 Hiromichi Nagao (The University of Tokyo)
Deep Learning to Extract Earthquakes and Low-Frequency Tremors in Continuous Seismic Waveforms
14:15 - 14:30 Takashi Shimokawabe (The University of Tokyo)
Fast Prediction Methods for Fluid Simulation Results Using Deep Neural Networks
Presentation Material
14:30 - 15:00 Richard Vuduc (Georgia Institute of Technology, USA)
(Keynote Talk) Data-movement accelerators for scientific computing problems
Presentation Material
15:00 - 15:20 (Discussion)
15:20 - 15:35 (Break)
15:35 - 15:50 Takeshi Ogita (Online) (Waseda University)
Verified Solutions of Large Sparse Linear Systems Arising from 3D Poisson Equation
Presentation Material
15:50 - 16:05 Masatoshi Kawai (Online) (Nagoya University)
Effectiveness of low/adaptive precision with ICCG method
Presentation Material
16:05 - 16:20 Hisashi Yashiro (Online) (National Institute for Environmental Studies)
Algorithmic transformation from physical models to data-driven models using the coupling library: a case of a climate model
Presentation Material
16:20 - 16:35 Shinji Sumimoto (The University of Tokyo)
h3-Open-SYS/WaitIO: A System-wide Heterogeneous Communication Library to Couple Multiple MPI programs
Presentation Material
16:35 - 16:50 Hayato Shiba (Online) (University of Hyogo)
Deep learning of simulated glassy dynamics
16:50 - 17:20 Xing Cai (Simula Research Laboratory, Norway)
(Keynote Talk) Towards high-performance unstructured-mesh computations
Presentation Material
17:20 - 17:40 (Discussion)
17:40 - 17:45 (Closing)

Video Recording

Seminar format

  • The seminars are open to everyone, not limited to users of this center.
  • Participation is free, but please be sure to register in advance using the form below.

Application form

The deadline for registration is 12:00 p.m. one day prior to the date of the seminar.
 * We will send you information about the Zoom meeting for the seminar by e-mail.
 * We will start sending the above e-mail about one week before the date of the seminar.

 [Application has been closed]

Inquiries about the ASE Seminar

Professor Kengo Nakajima, ASE Seminar Coordinator
E-mail: nakajima@cc.u-tokyo.ac.jp
(Please replace "@" with "@" before sending the e-mail.)