A Mixed-Precision Model for Matrix-Splitting Methods

DOI10.48550/arXiv.2610.06414MaRDI QIDQ7367840arXivFDO

Authors Neil Lindquist, Joseph M. Derlaga, David Appelhans, Chales W. Jackson

Publication date 5 October 2026

Published in arXiv

Tldr We propose a mixed-precision technique that reduces data movement in stationary iterative methods by using lower precision for intermediate computations while preserving double-precision accuracy for final results.

Full work available at URL https://doi.org/10.48550/arXiv.2610.06414

arXiv classification

cs.CEmath.NA

Cites work

• A Mixed-Precision Model for Matrix-Splitting Methods
• A New Approach to Probabilistic Rounding Error Analysis
• A simple strategy for varying the restart parameter in GMRES\((m)\)
• Accelerating the solution of linear systems by iterative refinement in three precisions
• Accuracy and Stability of Numerical Algorithms
• Adaptive Precision Block-Jacobi for High Performance Preconditioning in the Ginkgo Linear Algebra Software
• Algebraic error analysis for mixed-precision multigrid solvers
• Convergence of nested classical iterative methods for linear systems
• Improving the Performance of the GMRES Method using Mixed-Precision Techniques
• Mixed precision algorithms in numerical linear algebra
• Newton's method in mixed precision
• The superlinear convergence behaviour of GMRES
• The Tortoise and the Hare Restart GMRES
• The University of Florida sparse matrix collection
• Three-precision iterative refinement with parameter regularization and prediction for solving large sparse linear systems

Cited in

(1)

• A Mixed-Precision Model for Matrix-Splitting Methods

This page was built for publication: A Mixed-Precision Model for Matrix-Splitting Methods

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7367840)