Sort by
Refine Your Search
-
Listed
-
Category
-
Program
-
Employer
- Forschungszentrum Jülich
- Fraunhofer-Gesellschaft
- DAAD
- Technical University of Munich
- Free University of Berlin
- Alfred-Wegener-Institut Helmholtz-Zentrum für Polar- und Meeresforschung
- Leibniz
- Max Planck Institute for Demographic Research (MPIDR)
- Academic Europe
- GFZ Helmholtz-Zentrum für Geoforschung
- Heidelberg University
- Helmholtz-Zentrum Dresden-Rossendorf - HZDR - Helmholtz Association
- Max Planck Institute for Demographic Research, Rostock
- Max Planck School of Cognition
- Nature Careers
- TU Dresden
- Technische Universität Dortmund
- University of Bonn •
- 8 more »
- « less
-
Field
-
-edge Machine Learning applications on the Exascale computer JUPITER. Your work will include: Developing, implementing, and refining ML techniques suited for the largest scale Parallelizing model training
-
engineered 3D hydrogels, we will experimentally probe the mechanical forces and physical constraints that drive coordinated cell behavior. In parallel, we will develop and apply computational models and
-
Helmholtz-Zentrum Dresden-Rossendorf - HZDR - Helmholtz Association | Dresden, Sachsen | Germany | about 7 hours ago
to minimize training effort # Devise appropriate metrics to evaluate and tune trained models with respect to reproduction of key physical results # Contribute to a parallel training workflow to stream data from
-
on the Exascale computer JUPITER. Your work will include: Developing, implementing, and refining ML techniques suited for the largest scale Parallelizing model training and optimizing the execution User support in
-
cell types. Optimize 3D CAD designs for precision and parallel measurements. Evaluate the feasibility of integrating the probe system onto a robotic end-effector and design suitable mechanical and
-
build reliable, reproducible data flows for large EO datasets and workflows Lead performance engineering (parallelization, optimization, benchmarking) for adaptation and inference at scale Work closely
-
possible. What you will do You will be trained step by step in plant operation and actively support ongoing production processes. You will support the execution of quality controls in parallel with
-
and optimize large-scale training and inference runs for foundation models on JUPITER (multi-GPU/node, mixed precision, parallelization, I/O optimization) Integrate multimodal data sources (e.g., scRNA
-
sintering press with selected copper pastes, followed by detailed characterization of the resulting interfaces in terms of porosity, thermal and mechanical integrity. In parallel, simulation models will be
-
interested in the position, please submit your application by 17.12.2025 via our application portal and in parallel by e-mail to the President of the Fraunhofer-Gesellschaft, Professor Dr.-Ing. Holger Hanselka