Parallel Numerics - Winter 08
- Term
- Winter 08
- Lecturer
- Univ.-Prof. Dr. Thomas Huckle
- Time and Place
- Monday, 10:00-12:00, Room 02.07.023, First Lecture: t.b.a.
- Audience
- CSE (Compulsory Course), Mathematics (Master), Informatics (Master)
- Tutorials
- Friday, 10:00-12:00, Room 02.07.023, First Tutorial: t.b.a.
- Exam
- t.b.a.
- Semesterwochenstunden / ECTS Credits
- 2V + 2Ü SWS/ 5 Credits
- TUMonline
- {{{tumonline}}}
This course will be given in every winter term. The lectures and tutorials are conducted in English, and the course substitutes the German lecture "Numerik auf Parallelrechnern".
News
- Monday, October t.b.a.: 10:15-11:45 am: First lecture; Remarks on the tutorials
- Friday, October t.b.a.: 10:15-11:45 am: First tutorial
Contents
- High-Performance Computing
- Performance: Analysis, Modeling, and Measurements
- Basic Linear Algebra Subprograms
- Direct Solution of Sparse Linear Systems
- Iterative Methods for Linear Systems
- Linear Eigenvalue Problems
- Programming in MPI
Course Material
Lecture Notes
Slides
- Introduction (Winter 2007)
- Data Dependency Graphs (Winter 2007)
- BLAS (Winter 2007)
- Dense Matrices (Winter 2007)
- Sparse Matrices (Winter 2007)
- Preconditioners (Winter 2007)
- Domain Decomposition (Winter 2007)
- Computing Eigenvalues (Winter 2007)
- Iterative Methods - Stationary Methods, Krylov Methods (Winter 2006)
- Preconditioning - Stationary Preconditioners, Parallel Preconditioning, SPAI, Domain Decomposition (Winter 2006)
- Domain Decomposition (Winter 2006)
- Eigenvalues - QR, Jacobi, Divide & Conquer (Winter 2006)
- Discrete Fourier Transform (Winter 2006)
- Quantum Computing (Winter 2006)
Tutorials
- Flynn's Taxonomy & MPI Basics (Worksheet, Source Code)
- Numerical Integration & P2P Communication I (Worksheet, Source Code)
- Vector-Vector Operations & P2P Communication II (Worksheet, Source Code)
- Matrix / Matrix Operations & P2P Communication III (Worksheet, Source Code)
- Parallel Gauss Elimination & Collaborative Operations (Worksheet, Source Code)
- Tridiagonal Matrices, the Hockney/Golub Method & Message Tags (Worksheet)
- Sparse Matrix-Vector Multiplication & MPI Communicators (Worksheet)
- Stationary Methods (Worksheet, Source Code)
- Domain Decomposition (Worksheet)
Literature & External Links
- Numerical Linear Algebra for High-Performance Computers (Dongarra, Duff, Sorensen, van der Vorst)
- Parallel Algorithms for Matrix Computations (Gallivan, Heath, Ng, Ortega,...)
- A User's Guide to MPI (Pacheco)
- Iterative Methods for Sparse Linear Systems (Saad)
- Loesung linearer Gleichungssysteme auf Parallelrechnern (Frommer)
Exam
Details on both time and registration will be announced asap.