採択課題 【詳細】
| jh250043 | Innovative Multigrid Methods III |
|---|---|
| 課題代表者 | 藤井昭宏(工学院大学 情報学部) Akihiro Fujii ( Kogakuin University/School of informatics) |
| 概要 | We started the JHPCN project(Innovative Multigrid Methods) in FY.2018 with the aim of solving various technical problems in the parallel multigrid method and realizing basic research through international collaboration toward the exascale era. Parallel-in-Space/Time was added as a new research item with additional members including international ones. |
| 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） |
| [] Ryo Yoda, Matthias Bolten, ``Mixed-precision implementation of block ε-circulant preconditioner on modern CPU-GPU environments'', Japan Society for Industrial and Applied Mathematics (JSIAM) Letters, JSIAML-25R017 | |
| [] 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(2025). | |
| (2) 国際会議プロシーディングス （査読あり） | |
| [] Ryo Yoda, Matthias Bolten, ``Block Epsilon-Circulant Preconditioning with GPU-Accelerated Spatial Solvers for Linear Time-Dependent PDEs'', 2024 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), Milan, Italy, June 3, (2025) | |
| [] Nakajima, K., Furumura, T., Boillod-Cerneux, F., Di Napoli, E., Suarez, E., Arakawa, T., Sumimoto, S., Yashiro, H., Innovative Computational Science by Integration of Simulation/Data/Learning on Heterogeneous Supercomputers, June 10-13, ISC-HPC 2025 Project Poster, June | |
| (3) 国際会議発表（査読なし） | |
| [] Ryo Yoda, Matthias Bolten, ``Toward efficient solvers using block-epsilon circulant preconditioning on modern integrated CPU-GPU systems'', PinT 2025: the 14th Workshop on Parallel-in-time Integration, Parallel-in-time algorithms for exascale applications, ICMS, Bayes Centre, Edinburgh, UK, July 7--12, 2025. | |
| [] Ryo Yoda, Eric C. Cyr, Jacob Schroder, Matthias Bolten, Stephanie Friedhoff, ``Layer-Parallel Training with PyMGRIT for Deep Residual Neural Networks'', 22st Copper Mountain Multigrid Method Conference, Colorado, U.S., Sunday April 13--April 17, 2025. | |
| [] 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 | |
| (4) 国内会議発表（査読なし） | |
| [] 工藤侑也, 佐竹祐樹, 深谷猛, 岩下武史；「ESSC法を前処理とするCG法のGPU実装とその有効性検証」, The cross-disciplinary Workshop on Computing Systems, Infrastructures, and Programming (xSIG 2025), 於 高松市, 2025年8月. | |
| [] Matthias Bolten, Ryo Yoda, ``Parallelization of all-at-once preconditioned solvers for time-dependent PDEs'', Householder Symposium XXII, Ithaca, NY, USA, June 8--13, 2025 | |
| [] 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） | |
| [] 住元真司，胡曜，山崎一哉，中島研吾，WaitIO-Router:異種ネットワーク間高速通信を実現する通信ライブラリの試作（2025-HPC-200-29） | |
| (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） | |
| 該当なし |
無断転載禁止