Ruprecht-Karls-Universität Heidelberg
Siegel der Universität Heidelberg

Module for [Scientific Computing]

[back] to List of Modules.

[Parallel Solution of Large Linear Systems] - [2015 Sommer]

Module Code
IPLGG
Name
Parallel Solution of Large Linear Systems
Credit Points
8 LP CP
Workload
240 h
Duration
ein Semester
Cycle
0
Methods Lecture 4 h + Exercise course 2 h
Objectives
Content Finite element method for the discretization of elliptic partial differential equations Abstract subspace correction method Overlapping Schwartz method Geometric and algebraic multigrid method Non-overlapping domain decompostion method Convergence theory based on subspace correction method Implementation aspects and scalability of these methods
Learning outcomes Students are able to understand the structure and properties of linear systems arising from the discretization of elliptic partial differential equations formulate and analyse a variety of different subspace correction methods for the solution of these linear systems understand the algorithmic details of a software implementation of these methods in a message passing environment
Prerequisitesnone
Suggested previous knowledge Programming in C++, parallel computing, MA7, IPHR, ICC1, numerical solution of partial differential equations (MH7)
Assessments weekly excercises, programming projects, written or oral exam
Literature
zum Seitenanfang