Solving dense generalized eigenproblems on multi-threaded architectures

DOI10.1016/J.AMC.2012.05.020zbMATH Open1278.65037arXiv1111.6374OpenAlexW2125872426MaRDI QIDQ387520FDOQ387520

Authors: José I. Aliaga, Paolo Bientinesi, Davor Davidović, Edoardo Di Napoli, Francisco D. Igual, Enrique S. Quintana-Ortí 

Publication date: 23 December 2013

Published in: Applied Mathematics and Computation (Search for Journal in Brave)

Abstract: We compare two approaches to compute a portion of the spectrum of dense symmetric definite generalized eigenproblems: one is based on the reduction to tridiagonal form, and the other on the Krylov-subspace iteration. Two large-scale applications, arising in molecular dynamics and material science, are employed to investigate the contributions of the application, architecture, and parallelism of the method to the performance of the solvers. The experimental results on a state-of-the-art 8-core platform, equipped with a graphics processing unit (GPU), reveal that in real applications, iterative Krylov-subspace methods can be a competitive approach also for the solution of dense problems.

Full work available at URL: https://arxiv.org/abs/1111.6374

Recommendations

• The solution of large dense generalized eigenvalue problems on the Cray X-MP/24 with SSD
• scientific article; zbMATH DE number 434535
• Parallel implementations for solving generalized eigenvalue problems with symmetric sparse matrices
• scientific article; zbMATH DE number 1113885
• scientific article; zbMATH DE number 434520

zbMATH Keywords

eigenvaluesgraphics processorsKrylov-subspace iterationmulti-core architecturestridiagonal form

Mathematics Subject Classification ID

Parallel numerical computation (65Y05) Numerical computation of eigenvalues and eigenvectors of matrices (65F15)

Cites Work

• Title not available (Why is that?)
• High-performance solvers for dense Hermitian eigenproblems
• Algorithm 807
• Title not available (Why is that?)
• Rounding error analysis of the classical Gram-Schmidt orthogonalization process
• Programming matrix algorithms-by-blocks for thread-level parallelism
• Multiple representations to compute orthogonal eigenvectors of symmetric tridiagonal matrices
• Exploiting thread-level parallelism in the iterative solution of sparse linear systems
• A Parallel Eigensolver for Dense Symmetric Matrices Based on Multiple Relatively Robust Representations

Cited In (3)

• A fast band-Krylov eigensolver for macromolecular functional motion simulation on multicore architectures and graphics processors
• The geometric mean algorithm
• Exploring large macromolecular functional motions on clusters of multicore processors

Uses Software

• FLAME
• LAPACK
• Magma
• CUDA
• ARPACK
• libflame
• SBR Toolbox
• FLEUR
• iMod

This page was built for publication: Solving dense generalized eigenproblems on multi-threaded architectures

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q387520)