Available Master topics: Scientific Computing
How can DGX-2---the computing powerhouse containing 16 Nvidia V100 GPUs---be efficiently programmed and used for unstructured mesh computations?
This master project will use eX3---the Norwegian national infrastructure for experimental exploration of exascale computing---as a testbed for developing methodologies and tools that enable automated benchmarking and reporting of cutting-edge heterogeneous computing systems.
Almost all of today's computers, from smart phones to supercomputers, use multiple cores and thus parallel processing. This thesis looks at implementing challenging fundamental problems on today’s parallel computers.
How can we distribute complex problems on supercomputers that contain many CPU cores and GPUs ? This thesis looks at improving the software that decomposes problems for such parallel heterogeneous systems.
The goal for the project is to analyze and test a family of solvers for ordinary differential equations (ODEs) known as generalized Rush-Larsen (GRL) methods.