Empa - the place where innovation starts
Empa is the research institute for materials science and technology of the ETH Domain and conducts cutting-edge research for the benefit of industry and the well-being of society.
The Laboratory for Air Pollution / Environmental Technology has an opening for a
2-year postdoc/software engineer position on accelerating the atmospheric chemistry transport model ICON-ART for next-generation supercomputer architectures
ICON-ART is a new advanced chemistry transport model coupled to the numerical weather prediction model ICON (ICOsahedral Nonhydrostatic Model) that was developed primarily by the Deutscher Wetterdienst (DWD) and Max Planck Institute for Meteorology (MPI-M). The ART extension for Aerosols and Reactive Trace gases was developed at the Karlsruhe Institute of Technology (KIT) and is an official component of ICON.
In a new project under the Swiss Platform for Advanced Scientific Computing (PASC), we aim to develop a computationally highly efficient version of ICON-ART, optimized for the next generation supercomputing systems. For this purpose, large parts of ART will be ported to GPUs based on OpenACC directives. This will enable integrating ART into an already existing GPU-accelerated version of ICON.
We will closely collaborate with out project partners at KIT, ETH Zurich, the Swiss National Supercomputing Center, and with the EXCLAIM project at ETH. We will build on our extensive expertise, acquired during the porting of the previous numerical weather prediction model COSMO.
- Porting of selected ART components to GPUs using OpenACC directives.
- Setting up a testbed environment, evaluate and optimize the ported components.
- Identify I/O bottlenecks and optimize ICON-ART for I/O.
- Prepare and benchmark ICON-ART for selected target applications and demonstrate the new possibilities (e.g. in terms of resolution) offered by the efficient code.
- Participate in project meetings, manage Empa's tasks and deliverables in the project.
- MSc degree in physics, computational science, atmospheric sciences, computational fluid dynamics or a related discipline.
- Experience in similar projects and/or a PhD in a related topic.
- A strong background in computational modelling; experience in atmospheric modelling is an advantage.
- Excellent programming skills (e.g., Fortran, Python), experience with parallel and distributed computing (MPI, OpenMP) and with HPC systems; experience with GPU programming and OpenACC is an advantage.
- Ability to communicate scientific results in peer-reviewed journals and conferences.
- Fluency in English is required; basic knowledge of German is desirable.
The successful candidate will benefit from a highly stimulating research environment with strong national and international collaborations, a large team of experienced model developers and users, and excellent computing infrastructure. Working place will be Empa in Dübendorf.
Appointments will start preferably before 1 November 2021. All applications received before 15 September 2021 will receive full consideration.
For further details on the project HAMAM, ICON-ART, and the "sister project" EXLAIM please refer to: