3D Anisotropic Problem AMG Methods
收藏arXiv2025-09-30 收录
下载链接：
https://hpc.llnl.gov/hardware/compute-platforms/ruby
下载链接
链接失效反馈官方服务：
资源简介：
该数据集涵盖了针对100x100x100网格上的三维各向异性问题，使用标准7点拉普拉斯算子，优化后的代数多重网格（AMG）方法作为求解器和预处理器。此外，数据集还包括了针对不同问题变体和大小的优化求解器G3P-1和G3P-2的性能指标评估。该数据集的任务是针对三维各向异性问题的AMG方法的优化研究。
This dataset contains 3D anisotropic problems discretized on a 100×100×100 grid, which employ the standard 7-point Laplacian operator, with the optimized algebraic multigrid (AMG) method serving as both the solver and preconditioner. Furthermore, the dataset also includes performance metric evaluations of the optimized solvers G3P-1 and G3P-2 for various problem variants and scales. The core objective of this dataset is to support optimization research on AMG methods for 3D anisotropic problems.
提供机构：
Lawrence Livermore National Laboratory
