[IHASC] - [2022Sommer] - [en] - [Hardware Aware Scientific Computing]


Hardware Aware Scientific Computing [2022/23 WiSe]
Code
IHASC
Name
Hardware Aware Scientific Computing
CP
8
Duration
one semester
Offered
irregular
Format
Lecture 4 SWS + Exercise Course 2 SWS
Workload
240h;thereof
90h lecture
15h preparation for exam
135h self-study and working on assignments/projects (optionally in groups)
Availability
M.Sc. Angewandte Informatik
M.Sc. Data and Computer Science
M.Sc. Scientific Computing
Language
English
Lecturer(s)
Peter Bastian
Examination scheme
Learning objectives Students are familiar with different forms of parallelism in modern computer architectures. The can exploit this parallelism selecting an appropriate programming model. They are familiar with modelling of parallelism and know fundamental parallel algorithms from scientific computing.
Learning content Parallel Computer Architecture
- Pipelining and super-scalar processors, SIMD vectorisation
- Caches
- Multicore architectures
- GPUs
- Communication networks
Programming Models
- Shared memory programming with OpenMP and C++ threads
- OpenCL or Cuda
- Task-based programming
- Message-passing, MPI
Parallel Algorithms
- Speedup & scalability
- Roofline model
- Linear Algebra: Matrix-Vector, Matrix multiplication, solving dense
systems, solving sparse systems
- Iterative Solution of Linear Systems
- High-Performance Libraries
- Differential equations
- Particle Methods
Requirements for participation basic knowledge in computer architecture and numerical methods; good programming skills in C++
Requirements for the assignment of credits and final grade The module is completed with a graded examination. The final grade of the module is determined by the grade of the examination. Details for this examination as well as the requirements for the assignment of credits will be given by the lecturer an the beginning of this course.
Useful literature Frédéric Magoules, François-Xavier Roux, Guillaume Houzeaux: Parallel Scientific Computing, Wiley, 2016, doi: 10.1002/9781118761687