Sparse Linear Systems Solvers

Given a general square sparse matrix A and an RHS vector b, S3L_gen_iter_solve solves the linear system of equations Ax = b using an iterative algorithm, with or without preconditioning.

S3L_gen_iter_solve implements several different algorithms, which the programmer can select. These include

• BiConjugate Gradient Stabilized (BiCGSTAB)

• Conjugate Gradient (CG)

• Conjugate Gradient Squared (CGS)

• Conjugate Residuals (CR)

• Restarted Generalized Minimum Residual (GMRES)

• Quasi-Minimal Residual (QMR)

• Richardson method

If preconditioning is used, the programmer can choose point Jacobi or incomplete LU as the method. S3L_gen_iter_solve also allows the programmer to specify a number of other algorithm-specific parameters, including

• the convergence/divergence criteria

• the initial guess

• the maximum number of iterations