New Concept and Parallel Algorithms for Robust
Preconditioning in Large Scale Parallel Matrix Computation
Principal Investigator: Jun Zhang
Graduate Research Assistants: Yuan Hong
Graduate Research Assistants: Kai Wang (graduated with a Ph.D. in 2003)
Funding Sources: National Science Foundation
Funding Division: Advanced Computational Infrustrature and Reseach
Funding Program: Advanced Computational Research
Program Director: Xiaodong Zhang
Contract Number: ACI-0202934
Estimated Budget: $172,361
Duration: 06/01/2002 - 05/31/2005 (36 months)
Large sparse unstructured matrices arising from various
computer simulation and modeling are commonly solved
by preconditioned iterative methods. This research project
will study and design robust high performance
preconditioners for parallel solution of large sparse
linear systems, based on a class of multistep
successive sparse approximate inverse preconditioning
We will develop new concept and parallel
algorithms of multistep successive preconditioning
for enhancing the robustness of standard sparse
approximate inverse preconditioning techniques, and generalize
this concept to the context of other preconditioning
techniques. Study will
be conducted to show the advantages of such approach to
enhance both preconditioning accuracy and factorization
stability. We will build portable software packages to
preconditioning strategies for solving unstructured
general sparse linear systems on high performance parallel
The general purpose high performance preconditioned iterative
solvers from this research project are expected to make
significant impact in the field of applied scientific computing.
Our experience and existing strength will ensure that the
project be carried out fully as proposed. As U.S. industry
is more and more relying on computer aided design and
scale computer simulation and modeling will be a vital
component in new products research and development. The outcome
of this research will benefit U.S. industry as well as scientific
research community by providing more efficient kernel
software for large scale computer simulations.
Technical Reports and Computer Software:
MSP: a class of parallel multistep successive sparse approximate
inverse preconditioning strategies,
by Kai Wang and Jun Zhang. (January, 2002).
A class of parallel multilevel sparse approximate
inverse preconditionrs for sparse linear systems,
by Kai Wang, Jun Zhang, and Chi Shen. (November, 2002).
A comparative study on dynamic and static sparsity patterns
in parallel sparse approximate inverse preconditioning,
by Kai Wang, Sangbae Kim, and Jun Zhang. (November, 2002).
Parallel multistep successive sparse approximate inverse
preconditioning for solving general sparse linear systems
by Kai Wang, and Jun Zhang. (May 20, 2003).
Conference, Workshop, and Seminar Presentations:
A class of new parallel preconditioning strategies for solving large sparse
presented (by Jun Zhang) at the 2002 International Conference on Parallel
and Distributed Processing techniques and Applications, Las Vegas, NV,
June 24 - 27, 2002.
Parallel multilevel sparse approximate inverse preconditioner for solving
large sparse linear systems
presented (by Kai Wang, Ph.D. student) at the SIAM 2002 Annual Meeting,
Philadelphia, PA, July 8 - 12, 2002.
Robust parallel matrix preconditioning through successive sparse
Kai Wang and Jun Zhang,
presented (by Jun Zhang) at the 2nd International Workshop on Parallel
Matrix Algorithms and Applications, Neuchhatatel, Switzerland,
November 7 - 10, 2002.
Robust parallel preconditioning techniques for solving general sparse
Kai Wang and Jun Zhang,
presented (by Jun Zhang) at the 2002 International Symposium on
Distributed Computing and Applications to Business, Engineering and
Science, Wuxi, China, December 16 - 19, 2002.
Global and localized parallel preconditioning techniques for large
scale solid earth simulations,
Kai Wang, Sangbae Kim, Jun Zhang, Kengo Nakajima, and Hiroshi Okuda,
presented (by Jun Zhang) at the 2003 International Workshop on
Parallel and Distributed Scientific and Engineering Computing with
Applications, Nice, France, April 22-26, 2003.
Parallel Multilevel Sparse Approximate Inverse Preconditioners
in Large Sparse Matrix Computations,
Kai Wang, Jun Zhang, and Chi Shen, presented (by Jun Zhang) at
Supercomputing 2003 Igniting Innovation, Phoenix, AZ,
November 17-20, 2003.
This page is supported by the
U.S. National Science Foundation. However, any opinions, findings, and
conclusions or recommendations expressed in this documents are those of the
author and do not necessarily reflect the views of the
U.S. National Science Foundation.
Go back to
Funded Research Projects page.
This page was created on Wednesday, July 17, 2002, by
Last modified on Wednesday, November 26, 2003.