The position is a fixed-term position for a duration of 3 years. Appointment to the position of Postdoctoral Research Fellow is mainly intended to provide qualifications for work in top academic positions. It is a prerequisite that the applicant is able to carry out the project over the full course of the employment period.
The postdoc position is attached to the research project Efficient Machine Learning-Enhanced Modeling and Simulation on Exascale Architectures (EMLEMS) and will be involved in national project eX3 – Experimental Infrastructure for Exploration of Exascale Computing funded by the Research Council of Norway.
This postdoc project aims at improving the efficiency and performance of scientific modeling and simulations on exascale architectures through machine learning-enabled adaptivity, providing scalable, robust solutions with guaranteed accuracy in the least amount of time. Adaptivity is essential to increase the automation of the modeling and simulation workflow and to address the growing complexity of applications and architectures. Exascale architectures rely on the intricate interplay between thousands of heterogeneous processing nodes, each with a large number of cores, accelerators, memory types, and sophisticated interconnects. As a result, choosing optimal algorithms and implementations is highly application- and architecture-dependent. This project will investigate the opportunities of:
- leveraging scientific machine learning (SciML) in developing new algorithms, data layouts, and implementations that dynamically optimize the use of computational resources of exascale architectures, and
- leveraging high-performance scientific computing (HPC) in developing scalable and efficient machine learning training algorithms on exascale architectures.
Smart power systems with renewable energy will be an application domain in this project.
For details, please refer to the full announcement
