Post-doctoral Position: GRADES/Edge Computing/AI

Post-doctoral Position: GRADES/Edge Computing/AI

SOLEIL Synchrotron

Saint-Aubin, France

Context and missions

You will prototype an automatic diffraction data reduction application based on artificial intelligence algorithms.

This application will be developed on an "edge computing" device such as NVIDIA Jetson, then deployed on several beamlines (MARS, DIFFABS) in collaboration with the Detectors and Control-Acquisition groups.

The scope of the study will then be extended to small-angle data for the SWING beamline. Finally, a portability study of the approach to other beamlines involved in the PEPR DIADEME will be undertaken. You will participate in PEPR meetings and present the results obtained.

You will also contribute to the operational maintenance of GRADES' software, computing tools, and simulation tools.

Profile

We are looking for a candidate passionate about scientific computing, open-source software and systems (particularly Debian/Ubuntu). The candidate must hold a PhD in physics or scientific computing. Knowledge and/or experience in one or more of the following areas will be appreciated: data science, instrumentation, physics, AI, statistics, machine learning/deep learning.

Essential Knowledge and Skills:

• Mastery of scientific programming languages (Python, C/C++,…);
• Mastery of AI tools scikit-learn, scikit-image, tensorflow/pytorch;
• Mastery of development methods and tools (git, make/cmake);
• Basic knowledge of material diffraction (powder, small angles);
• Knowledge of open-source systems, particularly Debian/Ubuntu;
• Professional English: written, oral, technical.

Additional Knowledge:

• Scientific computing, applied mathematics, algorithmics;
• Physics, instrumentation;
• Scientific data format;
• Embedded GPU computing like NVIDIA Jetson;
• HPC programming (CUDA, OpenCL, MPI, OpenMP).

The mission also requires the following qualities:

• Teamwork and networking; Autonomy, priority management; Adaptability; Good interpersonal skills and service orientation; Being attentive to users and understanding their needs.