{"ID":2858344,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.08536","arxiv_id":"2510.08536","title":"Investigating Matrix Repartitioning to Address the Over- and Undersubscription Challenge for a GPU-based CFD Solver","abstract":"Modern high-performance computing (HPC) increasingly relies on GPUs, but integrating GPU acceleration into complex scientific frameworks like OpenFOAM remains a challenge. Existing approaches either fully refactor the codebase or use plugin-based GPU solvers, each facing trade-offs between performance and development effort. In this work, we address the limitations of plugin-based GPU acceleration in OpenFOAM by proposing a repartitioning strategy that better balances CPU matrix assembly and GPU-based linear solves. We present a detailed computational model, describe a novel matrix repartitioning and update procedure, and evaluate its performance on large-scale CFD simulations. Our results show that the proposed method significantly mitigates oversubscription issues, improving solver performance and resource utilization in heterogeneous CPU-GPU environments.","short_abstract":"Modern high-performance computing (HPC) increasingly relies on GPUs, but integrating GPU acceleration into complex scientific frameworks like OpenFOAM remains a challenge. Existing approaches either fully refactor the codebase or use plugin-based GPU solvers, each facing trade-offs between performance and development e...","url_abs":"https://arxiv.org/abs/2510.08536","url_pdf":"https://arxiv.org/pdf/2510.08536v1","authors":"[\"Gregor Olenik\",\"Marcel Koch\",\"Hartwig Anzt\"]","published":"2025-10-09T17:53:12Z","proceeding":"cs.DC","tasks":"[\"cs.DC\",\"cs.SE\"]","methods":"[]","has_code":false}