Mixed-precision conjugate gradient algorithm using the groupwise update strategy

DOI10.1007/S13160-024-00644-8MaRDI QIDQ6498423

Kensuke Aihara, Daichi Mukunoki, Katsuhisa Ozaki

Publication date: 7 May 2024

Published in: Japan Journal of Industrial and Applied Mathematics (Search for Journal in Brave)

Mathematics Subject Classification ID

Iterative numerical methods for linear systems (65F10)

Cites Work

• Unnamed Item
• Solving lattice QCD systems of equations using mixed precision solvers on GPUs
• Reliable updated residuals in hybrid Bi-CG methods
• Variants of residual smoothing with a small residual gap
• Performance and energy consumption of accurate and mixed-precision linear algebra kernels on GPUs
• A floating-point technique for extending the available precision
• Improved Error Bounds for Inner Products in Floating-Point Arithmetic
• The university of Florida sparse matrix collection
• Estimating the Attainable Accuracy of Recursively Computed Residual Methods
• Accuracy and Stability of Numerical Algorithms
• Accurate Sum and Dot Product
• Methods of conjugate gradients for solving linear systems
• Mixed precision algorithms in numerical linear algebra

This page was built for publication: Mixed-precision conjugate gradient algorithm using the groupwise update strategy