Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

RAPNet: Accelerating Algebraic Multigrid with Learned Sparse Corrections

About

The scalable solution of large sparse linear systems is a bottleneck in scientific computing and graph analysis. While algebraic multigrid (AMG) offers optimal linear scaling, its performance is severely constrained by the trade-off between the sparsity and convergence quality of coarse-grid operators. Classical AMG heuristics struggle to balance these objectives, often sacrificing stability or performance for sparsity. We propose RAPNet, a graph neural network (GNN) framework that resolves this trade-off by learning to generate sparse, robust coarse operators directly from the sparse algebraic system. Key to our approach is a level-wise training strategy that enables learning from small subgraphs and generalization to million-node domains, bypassing the bottlenecks of prior neural AMG attempts. RAPNet executes exclusively during the solver setup phase, ensuring that the solve phase retains its favorable computational properties. We show that our method outperforms classical non-Galerkin baselines on diverse PDE discretizations and graph Laplacians, making it particularly effective for multi-query tasks such as eigenproblems, time-dependent simulations, and inverse or design problems.

Yali Fink, Ido Ben-Yair, Lars Ruthotto, Eran Treister• 2026

Related benchmarks

TaskDatasetResultRank
Solving Sparse Linear Systems2D Geometric Symmetric
Average Iteration Count36
40
AMG Solver ConvergenceWatts Strogatz
Convergence Rate100
24
AMG Solver Convergence2D Geometric
Convergence Rate100
24
Solving Sparse Linear SystemsWatts-Strogatz Random Weights Asymmetric
Average Iteration Count109
23
Solving linear elasticity equations2D elasticity systems (λ = 10, μ = 1)
Iterations to Convergence55
20
AMG Solver Convergence3D FEM Advection-Diffusion
Convergence Rate100
18
Solving Sparse Linear SystemsSocial Hub Symmetric
Average Iterations9
18
Solving Sparse Linear Systems3D FEM Anisotropic Diffusion Symmetric
Average Iteration Count59
18
AMG Solver ConvergenceSocial Hub
Convergence Rate100
18
Solving Sparse Linear Systems3D Geometric Random Weights (Asymmetric)
Average Iteration Count30
18
Showing 10 of 15 rows

Other info

Follow for update