Distributed Schur Complement Techniques for General Sparse Linear Systems

  title={Distributed Schur Complement Techniques for General Sparse Linear Systems},
  author={Yousef Saad and Masha Sosonkina},
  journal={SIAM J. Sci. Comput.},
This paper presents a few preconditioning techniques for solving general sparse linear systems on distributed memory environments. These techniques utilize the Schur complement system for deriving the preconditioning matrix in a number of ways. Two of these preconditioners consist of an approximate solution process for the global system, which exploits approximate LU factorizations for diagonal blocks of the Schur complement. Another preconditioner uses a sparse approximate-inverse technique to… 

Figures from this paper

Parallel two level block ILU preconditioning techniques for solving large sparse linear systems
Robust parallel ILU preconditioning techniques for solving large sparse matrices
  • Chi ShenJun Zhang
  • Computer Science
    Proceedings 16th International Parallel and Distributed Processing Symposium
  • 2002
A parallel preconditioner for solving general sparse linear systems based on a two level block ILU factorization strategy is implemented and is fast and robust for solving certain large sparse matrices.
Parallel multilevel block ILU preconditioning techniques for large sparse linear systems
We present a class of parallel preconditioning strategies built on a multilevel block incomplete LU (ILU)factorization technique to solve large sparse linear systems on distributed memory parallel
Domain Decomposition and MultiLevel Type Techniques for General Sparse Linear Systems
A multi-level domain decomposition-type algorithm for iterative solution of large sparse linear systems based on independent subsets of nodes and a Schur complement technique that utilizes incomplete LU factorizations of local matrices are discussed.
Distributed block independent set algorithms and parallel multilevel ILU preconditioners
A Sparse Approximate Inverse Technique for ParallelPreconditioning of General Sparse Matrices
A sparse approximate inverse technique is introduced to solve general sparse linear systems and possesses much greater inherent parallelism than traditional preconditioners based on incomplete LU factorizations.
pARMS: A Package for Solving General Sparse Linear Systems on Parallel Computers
An overview of pARMS, a package for solving sparse linear systems on parallel platforms which integrates together variants from both Schwarz procedures and Schur complement-type techniques, is presented.
MSP: A Class of Parallel Multistep Successive Sparse Approximate Inverse Preconditioning Strategies
We develop a class of parallel multistep successive preconditioning strategies to enhance efficiency and robustness of standard sparse approximate inverse preconditioning techniques. The key idea is


Parallel Solution of General Sparse Linear Systems
This paper discusses a few algorithms and their implementations for solving distributed general sparse linear systems and compares a number of variants of Schwarz procedures with Schur complement techniques.
Approximate Inverse Techniques for Block-Partitioned Matrices
This paper proposes some preconditioning options when the system matrix is in block-partitioned form, which can help solve difficult linear systems whose coefficient matrices are highly indefinite.
Approximate inverse preconditioners for general sparse matrices
A number of preconditioners based on iterating on each column to get the approximate inverse of sparse-matrix by sparse-vector type operatoins are developed and compared on standard problems from the Harwell-Boeing collection.
Iterative solution of general sparse linear systems on clusters of workstations
This paper discusses issues and reports on the experience of PSPARSLIB, an on-going project for building a library of parallel iterative sparse matrix solvers, which aims to find efficient ways to precondition the system.
P-Sparslib: A Portable Library Of Distributed Memory Sparse Iterative Solvers
This paper explores issues related to data structures for distributed sparse matrices, simple matrix operations, as well as the implementations of the iterative solution kernels and analyze their performance on di erent architectures.
Approximate Inverse Preconditioners via Sparse-Sparse Iterations
Newton, "global," and column-oriented algorithms, and options for initial guesses, self-preconditioning, and dropping strategies are discussed, and some limited theoretical results on the properties and convergence of approximate inverses are derived.
Distributed ILU(0) and SOR Preconditioners for Unstructured Sparse Linear Systems
It is found that for the problems tested ILU(0) with multicoloring outperforms the other alternatives and the implementation of these two approaches on distributed memory computers and compare their performance on the CM-5 is considered.
A comparison of some domain decomposition and ILU preconditioned iterative methods for nonsymmetric elliptic problems
A version of the accelerated multiplicative Schwarz method is a consistently good performer and also the popular global ILU family of preconditioners on some nonsymmetric or inde nite elliptic model problems discretized by nite di erence methods.
A Flexible Inner-Outer Preconditioned GMRES Algorithm
  • Y. Saad
  • Computer Science
    SIAM J. Sci. Comput.
  • 1993
A variant of the GMRES algorithm is presented that allows changes in the preconditioning at every step, resulting in a result of the flexibility of the new variant that any iterative method can be used as a preconditionser.