Exploiting Lower Precision Arithmetic in Solving Symmetric Positive Definite Linear Systems and Least Squares Problems

DOI10.1137/19M1298263zbMath1467.65023OpenAlexW2997929922MaRDI QIDQ5857837

Nicholas J. Higham, Srikara Pranesh

Publication date: 8 April 2021

Published in: SIAM Journal on Scientific Computing (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1137/19m1298263

zbMATH Keywords

linear systempreconditioningconjugate gradient methodCholesky factorizationGMRESleast squares problemnormal equationssymmetric positive definite matrixiterative refinementdiagonal scalinghalf precision arithmetic

Mathematics Subject Classification ID

Iterative numerical methods for linear systems (65F10) Numerical computation of matrix norms, conditioning, scaling (65F35) Direct numerical methods for linear systems and matrix inversion (65F05) Preconditioners for iterative methods (65F08)

Related Items

Mixed precision algorithms in numerical linear algebra, Five-Precision GMRES-Based Iterative Refinement, Discretization-Error-Accurate Mixed-Precision Multigrid Solvers, Mixed-precision iterative refinement using tensor cores on GPUs to accelerate solution of linear systems, Numerical algorithms for high-performance computational science, Discretization-Error-Accurate Mixed-Precision Multigrid Solvers

Uses Software

• SparseMatrix
• mctoolbox
• CholQR
• CholeskyQR2

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Roundoff error analysis of the CholeskyQR2 algorithm
• Modified Cholesky algorithms: A catalog with new approaches
• Towards dense linear algebra for hybrid GPU accelerated manycore systems
• Convergence analysis of an algorithm for accurate inverse Cholesky factorization
• Condition numbers and equilibration of matrices
• The university of Florida sparse matrix collection
• Stability and Performance of Various Singular Value QR Implementations on Multicore CPU with a GPU
• A New Modified Cholesky Factorization
• Mixed Precision Block Fused Multiply-Add: Error Analysis and Application to GPU Tensor Cores
• Estimating the Attainable Accuracy of Recursively Computed Residual Methods
• A New Analysis of Iterative Refinement and Its Application to Accurate Solution of Ill-Conditioned Sparse Linear Systems
• Accelerating the Solution of Linear Systems by Iterative Refinement in Three Precisions
• Accuracy and Stability of Numerical Algorithms
• Optimal backward perturbation bounds for the linear least squares problem
• Mixed-precision iterative refinement using tensor cores on GPUs to accelerate solution of linear systems
• Shifted Cholesky QR for Computing the QR Factorization of Ill-Conditioned Matrices
• Squeezing a Matrix into Half Precision, with an Application to Solving Linear Systems
• A New Approach to Probabilistic Rounding Error Analysis
• Simulating Low Precision Floating-Point Arithmetic
• Mixed-Precision Cholesky QR Factorization and Its Case Studies on Multicore CPU with Multiple GPUs
• Accelerating Numerical Dense Linear Algebra Calculations with GPUs
• Three-Precision GMRES-Based Iterative Refinement for Least Squares Problems