Increasing the Performance of the Jacobi--Davidson Method by Blocking

DOI10.1137/140976017zbMath1329.65077OpenAlexW2175670603MaRDI QIDQ3457832

Gerhard Wellein, Georg Hager, Andreas Pieper, Melven Röhrig-Zöllner, Moritz Kreutzer, Jonas Thies, Achim Basermann, Andreas Alvermann, Holger Fehske

Publication date: 18 December 2015

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

Full work available at URL: https://semanticscholar.org/paper/8e8255c9bbbe7ac5a8ea9c0237f6a8417b22af76

zbMATH Keywords

algorithmsparse matrixnumerical experimenthigh performance computingblock methodsJacobi-Davidsonmulticore processorsperformance engineeringsparse eigenvalue problemshybrid parallel implementation

Mathematics Subject Classification ID

Computational methods for sparse matrices (65F50) Numerical computation of eigenvalues and eigenvectors of matrices (65F15) Parallel numerical computation (65Y05) Complexity and performance of numerical algorithms (65Y20)

Related Items (9)

Increasing the Performance of the Jacobi--Davidson Method by Blocking ⋮ Convergence of integration-based methods for the solution of standard and generalized Hermitian eigenvalue problems ⋮ Performance of the Low-Rank TT-SVD for Large Dense Tensors on Modern MultiCore CPUs ⋮ High-performance implementation of Chebyshev filter diagonalization for interior eigenvalue computations ⋮ Numerical bifurcation analysis of a 3D Turing-type reaction-diffusion model ⋮ The least squares and line search in extracting eigenpairs in Jacobi-Davidson method ⋮ Surface couplings for subdomain-wise isoviscous gradient based Stokes finite element discretizations ⋮ On short recurrence Krylov type methods for linear systems with many right-hand sides ⋮ Benefits from using mixed precision computations in the ELPA-AEO and ESSEX-II eigensolver projects

Uses Software

• JADAMILU
• BLAS
• MatrixMarket
• PRIMME
• FEAST
• Trilinos
• SparseMatrix
• ParMETIS
• PT-Scotch
• Sparsity
• SELL_C_sigma
• likwid
• STREAM
• TraceMIN
• STREAM benchmark

Cites Work

• Unnamed Item
• PRIMME
• On optimizing Jacobi-Davidson method for calculating eigenvalues in low dimensional structures using eight band \(\mathbf{k}\cdot\mathbf{p}\) model
• Parallel implementations of the trace minimization scheme trace{min} for the sparse symmetric eigenvalue problem
• JADAMILU: a software code for computing selected eigenvalues of large sparse symmetric matrices
• Das Verfahren der Treppeniteration und verwandte Verfahren zur Lösung algebraischer Eigenwertprobleme
• Inexact Newton preconditioning techniques for large symmetric eigenvalue problems
• A Grassmann--Rayleigh Quotient Iteration for Computing Invariant Subspaces
• Communication-optimal Parallel and Sequential QR and LU Factorizations
• FEAST As A Subspace Iteration Eigensolver Accelerated By Approximate Spectral Projection
• A Unified Sparse Matrix Data Format for Efficient General Sparse Matrix-Vector Multiplication on Modern Processors with Wide SIMD Units
• The university of Florida sparse matrix collection
• Numerical Methods for Large Eigenvalue Problems
• Increasing the Performance of the Jacobi--Davidson Method by Blocking
• An overview of the Trilinos project
• Nearly Optimal Preconditioned Methods for Hermitian Eigenproblems Under Limited Memory. Part II: Seeking Many Eigenvalues
• Cache efficient bidiagonalization using BLAS 2.5 operators
• The Jacobi-Davidson method
• Studies on Jacobi–Davidson, Rayleigh quotient iteration, inverse iteration generalized Davidson and Newton updates
• Jacobi--Davidson Style QR and QZ Algorithms for the Reduction of Matrix Pencils
• Convergence Analysis of Inexact Rayleigh Quotient Iteration
• Nearly Optimal Preconditioned Methods for Hermitian Eigenproblems under Limited Memory. Part I: Seeking One Eigenvalue
• An updated set of basic linear algebra subprograms (BLAS)
• On Improving Linear Solver Performance: A Block Variant of GMRES
• The One-Dimensional Hubbard Model
• An integrated Davidson and multigrid solution approach for very large scale symmetric eigenvalue problems

This page was built for publication: Increasing the Performance of the Jacobi--Davidson Method by Blocking