Hardware-oriented Krylov methods for high-performance computing

zbMATH Open1473.65003arXiv2104.02494MaRDI QIDQ5162633FDOQ5162633

Authors: Nils-Arne Dreier 

Publication date: 5 November 2021

Abstract: Krylov subspace methods are an essential building block in numerical simulation software. The efficient utilization of modern hardware is a challenging problem in the development of these methods. In this work, we develop Krylov subspace methods to solve linear systems with multiple right-hand sides, tailored to modern hardware in high-performance computing. To this end, we analyze an innovative block Krylov subspace framework that allows to balance the computational and data-transfer costs to the hardware. Based on the framework, we formulate commonly used Krylov methods. For the CG and BiCGStab methods, we introduce a novel stabilization approach as an alternative to a deflation strategy. This helps us to retain the block size, thus leading to a simpler and more efficient implementation. In addition, we optimize the methods further for distributed memory systems and the communication overhead. For the CG method, we analyze approaches to overlap the communication and computation and present multiple variants of the CG method, which differ in their communication properties. Furthermore, we present optimizations of the orthogonalization procedure in the GMRes method. Beside introducing a pipelined Gram-Schmidt variant that overlaps the global communication with the computation of inner products, we present a novel orthonormalization method based on the TSQR algorithm, which is communication-optimal and stable. For all optimized method, we present tests that show their superiority in a distributed setting.

Full work available at URL: https://arxiv.org/abs/2104.02494

Recommendations

• Strategies for the vectorized block conjugate gradients method
• Avoiding communication in nonsymmetric Lanczos-based Krylov subspace methods
• Auto-tuned Krylov methods on cluster of graphics processing unit
• On short recurrence Krylov type methods for linear systems with many right-hand sides
• Scalable linear solvers based on enlarged Krylov subspaces with dynamic reduction of search directions

Mathematics Subject Classification ID

Research exposition (monographs, survey articles) pertaining to computer science (68-02) Research exposition (monographs, survey articles) pertaining to numerical analysis (65-02) Iterative numerical methods for linear systems (65F10) Numerical algorithms for specific classes of architectures (65Y10)

Cited In (5)

• Avoiding communication in nonsymmetric Lanczos-based Krylov subspace methods
• Adaptively restarted block Krylov subspace methods with low-synchronization skeletons
• Strategies for the vectorized block conjugate gradients method
• General framework for deriving reproducible Krylov subspace algorithms: BiCGStab case
• Modified Krylov acceleration for parallel environments

This page was built for publication: Hardware-oriented Krylov methods for high-performance computing

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5162633)