学際大規模情報基盤共同利用・共同研究拠点

採択課題 【詳細】

jh250028 High resolution simulation of cardiac electrophysiology on realistic whole-heart geometries
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 Recent advancements in computational cardiology involve the use of cell-resolved mathematical models of electrophysiology, unlike traditional models that merge intra- and extracellular spaces. These models demand high computational resolution, necessitating advancements in high-performance computing (HPC). The project aims to extend the jh240030 project, enabling cell-resolved simulations of cardiac physiology to enhance understanding of the heart's electrical activities. The EMI (Extracellular-Membrane-Intracellular) model is a key focus, requiring specialized mesh partitioning and new numerical algorithms. The project also aims to leverage GPU computing, particularly with GH200 processors on the Miyabi system, to achieve unprecedented large-scale simulations, advancing knowledge in computational physiology and HPC.
関連Webページ
報告書等 研究紹介ポスター / 最終報告書
業績一覧 (1) 学術論文 (査読あり)
[] Kengo Nakajima, 2025, Optimization of communication-computation overlapping in parallel multigrid methods by process/thread allocation, Japan Journal of Industrial and Applied Mathematics, 42 (3), 1029-1062
[] Sergej Breiter, James D. Trotter, Karl Fürlinger, Cache partitioning for sparse matrix–vector multiplication on the A64FX, Parallel Computing, Vol. 127, 2026
(2) 国際会議プロシーディングス (査読あり)
[] James D. Trotter, Sinan Ekmekçibaşı, Doğan Sağbili, Johannes Langguth, Xing Cai, Didem Unat, 2025, CPU- and GPU-initiated Communication Strategies for Conjugate Gradient Methods on Large GPU Clusters, Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, 298-315
(3) 国際会議発表(査読なし)
[] Nakajima, K., Effects of Process/Thread Allocation for Optimization of Communication-Computation Overlapping in Parallel Multigrid Methods, 22nd Copper Mountain Multigrid Conference, Copper Mountain, CO, USA, April 15, 2025
[] Alessandro Gatti, James D. Trotter, Hermenegild Arevalo, Tor Skeie, Xing Cai, Towards a Resource-Efficient GPU Solver for Monodomain Equations in Cardiac Electrophysiology, Poster presented at The 52nd International Conference in Cardiology Computing (CinC), Sao Paulo, Brazil, September 14-17, 2025
[] 【Invited Talk】Nakajima, K., Long but “Straight” Road to FugakuNEXT via Miyabi, Connecting Global HPC Infrastructure with Grassroots Innovation, SCA/HPCAsia 2026, Osaka, Japan, January 27, 2026
[] 【Invited Talk】Nakajima, K., Integrations of Simulations/Data/Learning and Beyond on Miyabi, 3rd NVIDIA Supercomputing User Group (NVSUG) at SC25, November 20, 2025, St. Louis, MO, USA
[] 【Invited Talk】Nakajima, K., Our Next 10 Years with Supercomputing, Vision and Strategy: How will Supercomputing Centers contribute to the Future Development of HPC/AI+ ?, SCA/HPCAsia 2026, January 26-29, 2026, Osaka, Japan
(4) 国内会議発表(査読なし)
[] 【Invited Talk】中島研吾,スーパーコンピューティングへの招待:スパコンへの数学,第19回創造性の育成塾,東京,2025年7月29日
[] 【Invited Talk】中島研吾,三木洋平,山崎一哉,下川辺隆史,富岳NEXTにおける アプリケーションとプログラミング,RIMS共同研究:数値解析が拓く次世代情報社会~データ駆動型から「富岳NEXT」~,京都大学益川ホール,2025年10月8日~10日
[] 中島研吾,深谷猛,南里豪志,混合精度パイプライン型共役勾配法,情報処理学会研究報告(2025-HPC-199-02)
[] 中島研吾,深谷猛,南里豪志,大規模並列環境におけるパイプライン型共役勾配法,情報処理学会研究報告(2025-HPC-200-02)
[] 中島研吾,並列多重格子法によるおける通信と計算のオーバーラップ,第17回 自動チューニング技術の現状と応用に関するシンポジウム(ATTA2025),2025
[] 【Invited Talk】中島研吾,『計算・データ・学習』融合と富岳NEXT,その先にあるもの,第9回アドバンス・シミュレーションセミナー 2025年12月19日(オンライン)
(5) 公開したライブラリなど
該当なし
(6) その他(特許,プレスリリース,著書等)
[] 【Award】令和7年度文部科学大臣表彰科学技術賞科学技術振興部門:中島研吾,住元真司,荒川隆,下川邊隆史,塙敏博(Minister of MEXT Award for Science and Technology, Science and Technology Promotion Division, Reiwa 7:Kengo Nakajima, Shinji Sumimoto, Takashi Arakawa, Takashi Shimokawabe, Toshihiro Hanawa)
[] 【Award】山崎一哉,2025年度コンピュータサイエンス領域奨励賞(GPU化済の気象コードにおけるホストCPUコアの有効活用,情報処理学会研究報告2025-HPC-198-42,2025)
無断転載禁止