Block Locally Optimal Preconditioned Eigenvalue Xolvers (BLOPEX) in hypre and PETSc

  title={Block Locally Optimal Preconditioned Eigenvalue Xolvers (BLOPEX) in hypre and PETSc},
  author={Andrew V. Knyazev and Merico E. Argentati and Ilya Lashuk and Evgueni E. Ovtchinnikov},
  journal={SIAM J. Sci. Comput.},
We describe our software package Block Locally Optimal Preconditioned Eigenvalue Xolvers (BLOPEX) recently publicly released. BLOPEX is available as a stand-alone serial library, as an external package to PETSc (Portable, Extensible Toolkit for Scientific Computation, a general purpose suite of tools developed by Argonne National Laboratory for the scalable solution of partial differential equations and related problems), and is also built into hypre (High Performance Preconditioners, a… 

Figures and Tables from this paper

Recent implementations, applications, and extensions of the Locally Optimal Block Preconditioned Conjugate Gradient method (LOBPCG)

A review of the recent implementations and applications of LOBPCG, as well as extensions of the local optimality idea beyond standard eigenvalue problems.

C 2011 Society for Industrial and Applied Mathematics Nonsymmetric Multigrid Preconditioning for Conjugate Gradient Methods *

The main goal of this paper is to describe the possibility of turning off post-smoothing in geometric multigrid used as a preconditioner in conjugate gradient linear and eigenvalue solvers for the 3D Laplacian.

PFEAST: A High Performance Sparse Eigenvalue Solver Using Distributed-Memory Linear Solvers

This paper highlights a recent development within the software package that allows the dominant computational task, solving a set of complex linear systems, to be performed with a distributed memory solver.

On preconditioning for linear equations and eigenvalue problems

The main result is of a negative nature: it is shown that the conjugate gradient method with variable preconditioning may not yield any improvement relative to the preconditionsed steepest descent method.

Accelerating the LOBPCG method on GPUs using a blocked sparse matrix vector product

This paper presents a heterogeneous CPU-GPU implementation for a sparse iterative eigensolver -- the Locally Optimal Block Preconditioned Conjugate Gradient (LOBPCG), and proposes a GPU kernel based on a modified sliced ELLPACK format.

The Eigenvalues Slicing Library (EVSL): Algorithms, Implementation, and Software

This paper describes a software package called EVSL (for EigenValues Slicing Library) for solving large sparse real symmetric standard and generalized eigenvalue problems. As its name indicates, the

Efficient parallel solution to large‐size sparse eigenproblems with block FSAI preconditioning

BFSAI coupled with a block diagonal incomplete decomposition proves a robust and efficient parallel preconditioner in a number of test cases arising from the finite element discretization of 3D fluid‐dynamical and mechanical engineering applications, outperforming FSAI even by a factor of 8 and exhibiting a satisfactory scalability.

Accelerating Certifiable Estimation with Preconditioned Eigensolvers

  • David M. Rosen
  • Computer Science
    IEEE Robotics and Automation Letters
  • 2022
This letter designs a specialized solution verification algorithm based upon the locally optimal block preconditioned conjugate gradient (LOBPCG) method together with a simple yet highly effective algebraic preconditionser, which is very effective in practice.



Preconditioned Eigensolver LOBPCG in hypre and PETSc

The current state of the LOBPCG software for hypre and PETSc is described and scalability results on distributed memory parallel clusters using domain decomposition and multigrid preconditioning are demonstrated.

Toward the Optimal Preconditioned Eigensolver: Locally Optimal Block Preconditioned Conjugate Gradient Method

  • A. Knyazev
  • Computer Science
    SIAM J. Sci. Comput.
  • 2001
Numerical results establish that the LOBPCG method is practically as efficient as the ``ideal'' algorithm when the same preconditioner is used in both methods, and direct numerical comparisons with the Jacobi--Davidson method show that the method is more robust and converges almost two times faster.

The Design and Implementation of hypre, a Library of Parallel High Performance Preconditioners

This chapter discusses the design of the conceptual interfaces in hypre and illustrates their use with various examples, and discusses the data structures and parallel implementation of these interfaces.

SLEPc: A scalable and flexible toolkit for the solution of eigenvalue problems

The Scalable Library for Eigenvalue Problem Computations (SLEPc) is a software library for computing a few eigenvalues and associated eigenvectors of a large sparse matrix or matrix pencil. It has

Preconditioned Eigensolvers: Practical Algorithms

Numerical comparison establishes that this method is much faster than the Block Steepest Ascent Method when the same preconditioner is used in both methods, and numerical results showing that the method converges linearly with the optimal convergence rate are presented.

Pursuing scalability for hypre's conceptual interfaces

The data structures, parallel implementation, and resulting performance of the IJ, Struct and semiStruct interfaces are described, which investigates their scalability, presents successes as well as pitfalls of some of the approaches and suggests ways of dealing with them.

A geometric theory for preconditioned inverse iteration. I : Extrema of the Rayleigh quotient

Preconditioned Eigensolvers - an Oxymoron?

The author tries to mention the majority of important theoretical results and ideas from the Soviet literature on the subject, adding references from the other side of the Iron Curtain usually just to preserve the integrity of the topic.

A comparison of eigensolvers for large‐scale 3D modal analysis using AMG‐preconditioned iterative methods

The goal of this paper is to compare a number of algorithms for computing a large number of eigenvectors of the generalized symmetric eigenvalue problem arising from a modal analysis of elastic structures by considering the use of preconditioned iterative methods.