A flexible CUDA LU-based solver for small, batched linear systems

DOI10.1007/978-3-319-06548-9_5MaRDI QIDQ5261399zbMATH OpenOpenAlexFDO

Authors Antonino Tumeo, Nitin Gawande, Oreste Villa

Publication date 3 July 2015

Published in Numerical Computations with GPUs (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/978-3-319-06548-9_5

zbMATH Keywords

performancenumerical exampleparallel computationCUDALU factorizationNewton-Raphson techniquesets of nonlinear equations

Mathematics Subject Classification ID

Direct numerical methods for linear systems and matrix inversion (65F05) Parallel numerical computation (65Y05) Numerical computation of solutions to systems of equations (65H10) Numerical algorithms for specific classes of architectures (65Y10)

Recommendations

• Efficient batch LU and QR decomposition on GPU
• Linear systems solvers for distributed-memory machines with GPU accelerators
• Batched triangular dense linear algebra kernels for very small matrix sizes on GPUs
• Accelerating preconditioned iterative linear solvers on GPU
• A guide for implementing tridiagonal solvers on GPUs

Describes a project that uses

Uses Software

• TOUGH2-MP
• PFLOTRAN
• CUBLAS
• CUDA

This page was built for publication: A flexible CUDA LU-based solver for small, batched linear systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5261399)