Conjugate gradients on multiple GPUs

DOI10.1002/FLD.2462zbMATH Open1206.65131OpenAlexW2020526771MaRDI QIDQ3068356FDOQ3068356

Authors Serban Georgescu, H. Okuda

Publication date 14 January 2011

Published in International Journal for Numerical Methods in Fluids (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1002/fld.2462

zbMATH Keywords

numerical examplesconvergence accelerationcondition numbersGPGPUconjugate gradientsPoisson's equationiterative refinementmixed precisioncommunication-computation overlappingsparse-matrix vector multiplication

Mathematics Subject Classification ID

Computational methods for sparse matrices (65F50) Iterative numerical methods for linear systems (65F10) Numerical computation of matrix norms, conditioning, scaling (65F35)

Recommendations

• Parallel preconditioned conjugate gradient algorithm on GPU
• Accelerating iterative linear solvers using multiple graphical processing units
• Accelerating preconditioned iterative linear solvers on GPU
• A factored sparse approximate inverse preconditioned conjugate gradient solver on graphics processing units
• MPI-CUDA sparse matrix-vector multiplication for the conjugate gradient method with an approximate inverse preconditioner

Cites work

• scientific article; zbMATH DE number 1953444 (Why is no real title available?)
• Deflated preconditioned conjugate gradient solvers for the pressure-Poisson equation
• Methods of conjugate gradients for solving linear systems
• Numerical Linear Algebra for High-Performance Computers
• Performance and accuracy of hardware-oriented native-, emulated- and mixed-precision solvers in FEM simulations

Cited in

(7)

• GPU acceleration for FEM-based structural analysis
• Multitasking the conjugate gradient method on the CRAY X-MP/48
• A GPU solver for symmetric positive-definite matrices vs. traditional codes
• Matrix-free GPU implementation of a preconditioned conjugate gradient solver for anisotropic elliptic PDEs
• scientific article; zbMATH DE number 7552283 (Why is no real title available?)
• An improved mixed Lagrangian-Eulerian (IMLE) method for modelling incompressible Navier-Stokes flows with CUDA programming on multi-gpus
• ``Seen is solution a CAD/CAE integrated parallel reanalysis design system

Describes a project that uses

Uses Software

• CUDA

This page was built for publication: Conjugate gradients on multiple GPUs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3068356)