| authors |
|
| name |
Philipp Krah |
| email |
philipp.krah@cea.fr |
| title |
Dr. Philipp Krah |
| role |
AI research engineer |
| orcid |
0000-0001-8982-4230 |
| user_groups |
|
| organizations |
|
| social |
| icon |
icon_pack |
link |
envelope |
fas |
mailto:philipp.krah@cea.fr |
|
|
Philipp Krah is a computational physicist with expertise in surrogate modeling, artificial intelligence (AI), and high-performance computing (HPC). He earned a Ph.D. in Applied Mathematics from TU Berlin, where he developed data-driven reduced-order models using AI and nonlinear reduction methods.
His research focuses on incorporating physical information into data-driven approaches to improve their realibility and performance. He has published in leading mathematical and physical journals and has contributed to several open-source HPC projects, including WABBIT (Fortran, OpenMPI), CMM-turbulence (C++, CUDA), and GyselaX++ (C++, Kokkos). Currently, he is working on non-axisymmetric gyrokinetic plasma simulations for stellarators and integrating AI-based diagnostics and reduced-order models into GyselaX++ as part of the NumPEx Exa-DoST program.