Performance and Accuracy of LAPACK's Symmetric Tridiagonal Eigensolvers

DOI10.1137/070688778zbMath1165.65014MaRDI QIDQ3630382

Beresford N. Parlett, Christof Vömel, Osni A. Marques, James W. Demmel

Publication date: 28 May 2009

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

Full work available at URL: https://digital.library.unt.edu/ark:/67531/metadc897405/

zbMATH Keywords

performance; algorithm; numerical examples; accuracy; tridiagonal matrix; inverse iteration; benchmark; symmetric eigenvalue problem; QR algorithm; bisection; divide and conquer; LAPACK; multiple relatively robust representations

Mathematics Subject Classification ID

65F50: Computational methods for sparse matrices

15A23: Factorization of matrices

65F15: Numerical computation of eigenvalues and eigenvectors of matrices

15A18: Eigenvalues, singular values, and eigenvectors

65Y20: Complexity and performance of numerical algorithms

Related Items

Communication lower bounds and optimal algorithms for numerical linear algebra, A note on generating finer‐grain parallelism in a representation tree, Restructuring the Tridiagonal and Bidiagonal QR Algorithms for Performance, Fast Computation of Spectral Projectors of Banded Matrices, Linear algebra software for large-scale accelerated multicore computing, Computing the Bidiagonal SVD Through an Associated Tridiagonal Eigenproblem, Performance and accuracy of the basic closure algorithm of quadrature-based moment methods, A limited memory BFGS algorithm for non-convex minimization with applications in matrix largest eigenvalue problem, Projection methods for quantum channel construction, Coupling bounds for approximating birth-death processes by truncation, On tridiagonal matrices unitarily equivalent to normal matrices, Higher-order spectral clustering for geometric graphs, Performance-Based Numerical Solver Selection in the Lighthouse Framework, Preconditioners for Saddle Point Systems with Trace Constraints Coupling 2D and 1D Domains, An Accelerated Divide-and-Conquer Algorithm for the Bidiagonal SVD Problem, New fast divide-and-conquer algorithms for the symmetric tridiagonal eigenvalue problem

Uses Software

• LAPACK
• BLAS
• JDQR
• SparseMatrix
• NWChem
• JDQZ