Object-oriented design of preconditioned iterative methods in diffpack

Bruaset, Are Magnus; Langtangen, Hans Petter

doi:10.1145/244768.244776

Cited by 34 publications

(15 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They describe a C++ package for that allows implementation of block preconditioners in a way that is independent of the storage scheme for the submatrix blocks. The Diffpack code of Bruaset and Langtangen [1997] is a C++ object-oriented implementation of iterative solvers for sparse linear systems.…”

Section: Related Workmentioning

confidence: 99%

Object-oriented software for quadratic programming

Gertz

Wright

2003

ACM Trans. Math. Softw.

209

132

View full text Add to dashboard Cite

The object-oriented software package OOQP for solving convex quadratic programming problems (QP) is described. The primal-dual interior point algorithms supplied by OOQP are implemented in a way that is largely independent of the problem structure. Users may exploit problem structure by supplying linear algebra, problem data, and variable classes that are customized to their particular applications. The OOQP distribution contains default implementations that solve several important QP problem types, including general sparse and dense QPs, bound-constrained QPs, and QPs arising from support vector machines and Huber regression. The implementations supplied with the OOQP distribution are based on such well known linear algebra packages as MA27/57, LAPACK, and PETSc. OOQP demonstrates the usefulness of object-oriented design in optimization software development, and establishes standards that can be followed in the design of software packages for other classes of optimization problems. A number of the classes in OOQP may also be reusable directly in other codes.

show abstract

Section: Related Workmentioning

confidence: 99%

Object-oriented software for quadratic programming

Gertz

Wright

2003

ACM Trans. Math. Softw.

209

132

View full text Add to dashboard Cite

show abstract

“…Considering numerical solution of partial differential equations by finite difference, volume, or element methods on structured grids, the most important type of operation from a performance point of view is the action of a stencil on a grid function. This action occurs in explicit finite difference schemes, matrix-vector products in Krylov subspace methods [7] useful for implicit schemes, as well as restriction, prolongation, and smoothers in multigrid solvers or preconditioners [6]. Even in complicated multi-physics codes, applying a stencil on the grid is usually the primary performance bottleneck.…”

Section: Experiments and Measurementsmentioning

confidence: 99%

On the Performance of the Python Programming Language for Serial and Parallel Scientific Computations

Cai

Langtangen

Moe

2005

Scientific Programming

Self Cite

View full text Add to dashboard Cite

This article addresses the performance of scientific applications that use the Python programming language. First, we investigate several techniques for improving the computational efficiency of serial Python codes. Then, we discuss the basic programming techniques in Python for parallelizing serial scientific applications. It is shown that an efficient implementation of the array-related operations is essential for achieving good parallel performance, as for the serial case. Once the array-related operations are efficiently implemented, probably using a mixed-language implementation, good serial and parallel performance become achievable. This is confirmed by a set of numerical experiments. Python is also shown to be well suited for writing high-level parallel programs.

show abstract

“…Generally, the better-conditioned system leads to an accelerated convergence in the iterative solution [4]. Some well-documented preconditioning methods such as incomplete LU factorization (ILU) and polynomial preconditioning methods [5]- [6] can be effective. However, they usually require well-above O(N ) operations to implement.…”

Section: Introductionmentioning

confidence: 99%

Accelerating the transient simulation of semiconductor devices using filter-bank transforms

Movahhedi

Abdipour

Dehghan

2005

Int. J. Numer. Model.

View full text Add to dashboard Cite

Abstract-Using fully or semi-implicit schemes to solve hydrodynamic transport equations is very suitable and efficient for transient simulations of semiconductor devices. But, using these techniques leads to a large system of linear equations. Here for the first time, a preconditioning method based on the filter-bank and wavelet transforms is used to facilitate the iterative solution of this system. As the first step in the performance investigation of this preconditioner in the simulation of semiconductor devices, we apply it to the modified Poisson's equation in Drift-Diffusion model. Numerical results show that the condition number are significantly reduced and convergence rate is increased. The most important advantage of this preconditioner rather than the other preconditioners is its low computational complexity which can be reduced to O(N ).

show abstract

Object-oriented design of preconditioned iterative methods in diffpack

Cited by 34 publications

References 13 publications

Object-oriented software for quadratic programming

Object-oriented software for quadratic programming

On the Performance of the Python Programming Language for Serial and Parallel Scientific Computations

Accelerating the transient simulation of semiconductor devices using filter-bank transforms

Contact Info

Product

Resources

About