swMATH ID: 11521
Software Authors: E. Chow
Description: ParaSails is a parallel sparse approximate inverse preconditioner for the iterative solution of large, sparse systems of linear equations. It is a self-contained module in the HYPRE preconditioner library currently being developed at the Center for Applied Scientific Computing. ParaSails has been used to solve finite element elasticity problems inside an LLNL simulation code with more than 4 million equations on 1000 processors of ASCI Blue-Pacific (IBM SP). It has also been demonstrated on anisotropic diffusion problems with 216 million equations. ParaSails uses least-squares (Frobenius norm) minimization to compute a sparse approximate inverse. The sparsity pattern used is the pattern of a power of a sparsified matrix. ParaSails also uses a post-filtering technique to reduce the cost of applying the preconditioner. The pattern of the preconditioner can be reused to generate preconditioners for different matrices in a sequence of linear solves. ParaSails solves symmetric positive definite (SPD) problems using a factorized SPD preconditioner. ParaSails can also solve general (nonsymmetric and/or indefinite) problems with a nonfactorized preconditioner. The software available to be downloaded includes parallel CG and GMRES solvers, a parallel matrix class and a test driver.
Homepage: https://computation.llnl.gov/casc/parasails/
Related Software: SparseMatrix; ILUT; PARDISO; ILUPACK; hypre; PETSc; MSPAI; FSAIPACK; GenASPI; pARMS; ARMS; CSparse; WSMP; BILUTM; LAPACK; JDQZ; BiCGstab; JDQR; BLAS; symrcm
Cited in: 30 Publications
Further Publications: https://computation.llnl.gov/casc/parasails/pubs.html

Citations by Year