Parallel unsmoothed aggregation algebraic multigrid algorithms on GPUs

DOI10.1007/978-1-4614-7172-1_5zbMATH Open1275.65084arXiv1302.2547OpenAlexW1790422192MaRDI QIDQ2847764FDOQ2847764

Authors: J. Brannick, Yao Chen, X. Hu, Ludmil T. Zikatanov 

Publication date: 11 September 2013

Published in: Numerical Solution of Partial Differential Equations: Theory, Algorithms, and Their Applications (Search for Journal in Brave)

Abstract: We design and implement a parallel algebraic multigrid method for isotropic graph Laplacian problems on multicore Graphical Processing Units (GPUs). The proposed AMG method is based on the aggregation framework. The setup phase of the algorithm uses a parallel maximal independent set algorithm in forming aggregates and the resulting coarse level hierarchy is then used in a K-cycle iteration solve phase with a $el{l}^1$-Jacobi smoother. Numerical tests of a parallel implementation of the method for graphics processors are presented to demonstrate its effectiveness.

Full work available at URL: https://arxiv.org/abs/1302.2547

Recommendations

• A parallel auxiliary grid algebraic multigrid method for graphic processing units
• Accelerating algebraic multigrid solvers on NVIDIA GPUs
• A GPU accelerated aggregation algebraic multigrid method
• Porting an aggregation-based algebraic multigrid method to GPUs
• Exposing fine-grained parallelism in algebraic multigrid methods

zbMATH Keywords

numerical examplesparallel computationmultigrid methodsgraphical processing units\(K\)-cycle iterationadaptive aggregationisotropic graph Laplacian problemsunsmoothed aggregation

Mathematics Subject Classification ID

Computational methods for sparse matrices (65F50) Parallel numerical computation (65Y05) Multigrid methods; domain decomposition for boundary value problems involving PDEs (65N55) Iterative numerical methods for linear systems (65F10) Laplace operator, Helmholtz equation (reduced wave equation), Poisson equation (35J05)

Cited In (21)

• AMG Preconditioners for Linear Solvers towards Extreme Scale
• Exposing fine-grained parallelism in algebraic multigrid methods
• Accelerating algebraic multigrid solvers on NVIDIA GPUs
• Algebraic multigrid for directed graph Laplacian linear systems (NS-LAMG).
• Algebraic Multigrid Using a Stencil–CSR Hybrid Format on GPUs
• Efficient Algebraic Multigrid Preconditioners on Clusters of GPUs
• Numerical study of geometric multigrid methods on CPU-GPU heterogeneous computers
• A nearly optimal multigrid method for general unstructured grids
• A GPU accelerated aggregation algebraic multigrid method
• A parallel auxiliary grid algebraic multigrid method for graphic processing units
• Development of a parallel algorithm based on an implicit scheme for the discontinuous Galerkin method for solving diffusion type equations
• An adaptive multigrid method based on path cover
• Parallel algebraic multigrid method for cable equation
• Toward Cost-Effective Reservoir Simulation Solvers on GPUs
• Matrix-free GPU implementation of a preconditioned conjugate gradient solver for anisotropic elliptic PDEs
• Accelerating iterative linear solvers using multiple graphical processing units
• Porting an aggregation-based algebraic multigrid method to GPUs
• Lean algebraic multigrid (LAMG): fast graph Laplacian linear solver
• AmgX: a library for GPU accelerated algebraic multigrid and preconditioned iterative methods
• Remarks on the numerical approximation of Dirac delta functions
• Preconditioned GMRES solver on multiple-GPU architecture

Uses Software

• BoomerAMG
• CUDA
• CUSP

This page was built for publication: Parallel unsmoothed aggregation algebraic multigrid algorithms on GPUs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2847764)