Programming and Parallel Scientific Computing
Syllabus
The course provides an introduction to C++ programming and algorithms with a focus on scientific computing using Cimlib_CFD. An overview of the C++ language is provided: specification, arithmetics, memory, performance, object-oriented design for component-based software, and advanced topics in numerics using templates. Distributed and shared-memory parallel computing are then approached with exercises related to numerical methods for PDEs.
Schedule
Location: I/E102 + Zoom Visioconference.
Time: AM: 09:00-12:00, PM: 13:30-16:30
| Week 46 | ||
| 15. November | I/E102 | AM: Introduction to supercomputing |
| I/E102 | PM: Finite-precision arithmetics and data models | |
| 17. November | I/E102 | AM: Procedural language programming: application to linear algebra |
| I/E102 | PM: Object-Oriented abstraction for numerical frameworks in C++: example of mesh topologies | |
| Week 47 | ||
| 22. November | I/E102 | AM: Models for parallel computing and performance metrics |
| 24. November | I/E102 | AM: Distributed Memory Programming Models |
| I/E102 | PM: Distributed Memory Programming Models | |
| Week 48 | ||
| 29. November | I/E102 | AM: Applications of distributed scientific computing (Jonathan Viquerat) |
| I/E102 | PM: Shared Memory Programming Models | |
| 01. December | I/E102 | AM: Iterative solvers and applications |
Project deadlines: TBA.