採択課題 【詳細】
| jh260046 | Innovative Computational Science by Integration of Simulation/Data/Learning on Heterogeneous Environments towards Fugaku NEXT |
|---|---|
| 課題代表者 | 中島研吾(東京大学情報基盤センター) Kengo Nakajima (Information Technology Center, The University of Tokyo) |
| 概要 | This project has been advancing computational science by integrating "Simulation/Data/Learning (S+D+L)” on heterogeneous supercomputers such as “Wisteria/BDEC-01 (U.Tokyo)” and “Miyabi (JCAHPC)”, along with "mdx" since FY,2021 with international partners from Germany, France, Croatia and USA using the innovative software platform “h3-Open-BDEC (https://h3-open-bdec.cc.u-tokyo.ac.jp/)”. In 2025, RIKEN announced its plan for Fugaku’s successor (Fugaku NEXT), which will start its operation after 2029 and will adopt Fujitsu’s Monaka-X with NVIDIA GPUs. The NVIDIA GH200 installed in Miyabi-G adopts NVLink-C2C, enabling efficient utilization of CPU and GPU, and similar technology is expected to be adopted in Fugaku NEXT. In FY.2026, with Fugaku NEXT in mind, we will make further promotion of Integration of (S+D+L) while developing applications in earth sciences, life sciences, and quantum sciences that efficiently coordinate GPU-to-GPU interactions on Miyabi-G. In FY.2025, we achieved the integration of Miyabi and a quantum computer (ibm_kawasaki). In FY.2026, we aim to extend this knowledge to interconnections between supercomputer systems and build a more flexible environment for Integration of (S+D+L). |
| 関連Webページ | |
| 報告書等 | 研究紹介ポスター ／ 最終報告書 |
無断転載禁止