This Introduces the three fundamental types of PDE's -- Elliptic, Parabolic and Hyperbolic and studies the numerical solution of Elliptic Equations |
The sparse matrix formulation is used and iterative approaches -- Jacobi, Gauss Seidel and SOR are defined |
These are motivated by analogies between equilibrium of diffusive equations and elliptic systems |
Parallel Computing is Discussed for Gauss Seidel |
Eigenvalue analysis is used to discuss convergence of methods |
We discuss Multigrid methods at a simple level |
001 CPS 615 -- Computational Science in Simulation Track Solution of Simple Partial Differential Equations and Iterative Solvers 002 Abstract of Simple Partial Differential Equations and Iterative Solvers 003 Boundary Conditions I 004 Boundary Conditions II 005 Boundary Conditions III 006 Iterative Methods for Solving Sparse Matrices 007 Formalism for Iterative Methods 008 Preconditioning 009 Convergence of Jacobi in One Dimension 010 What is Easy/Hard for Jacobi? 011 Information Moves Slowly 012 Lowest Eigenvalue converges fastest 013 Successive Over Relaxation I 014 Successive Over Relaxation II 015 Some Mathematical Details 016 Multigrid Methods 017 Gauss Seidel is Slow I 018 Gauss Seidel is Slow II 019 Multigrid Philosophically 020 Multigrid Hierarchy 021 Basic Multigrid Ideas 022 Multigrid Algorithm: procedure MG(level, A, u, f) 023 Multigrid Cycles 024 What can we do for Parallel Gauss Seidel? 025 16 by 16 Wavefront Parallel Gauss Seidel 026 Wavefront 027 Red Black Parallel Gauss Seidel I 028 Red Black 029 Red Black Parallel Gauss Seidel II 030 Parallel Red Black 031 Red Black Parallel Gauss Seidel III 032 Red Black Parallel Gauss Seidel IV