A graph neural network enhances algebraic multigrid solvers by predicting optimal polynomial coefficients for sparse pseudo-inverse operators. The method reduces V-cycle iterations and achieves wall-clock speedups of 4% to 37% across benchmarks, with robust performance on meshes up to 128 times larger than training data and on unseen industry problems like AirfRANS.
Graph Neural Networks Accelerate Algebraic Multigrid Pressure Solver
from English