Assistant / engineer position in neuromorphic engineering

We are looking for a motivated research assistant / engineer (“ingénieur d’étude” – IE) with expertise in neuromorphic engineering to join the team of Drs. Timothée Levi, Fabien Wagner, and Amélie Aussel at the University of Bordeaux (Institut du Matériau au Système – IMS – and Institut des Maladies Neurodégénératives – IMN).

The goal of the project is to expand our current efforts towards performing large-scale simulations of conductance-based neuronal models on FPGAs, with an application to neurostimulation of the  hippocampal formation. The initial contract would be for a period of 1 year with an expected starting date on Oct 1st, 2024. If interested, please check the information below and contact us at the indicated email address.

Context

At the Institute of Neurodegenerative Diseases (IMN), the teams of Drs. Wagner and Aussel have developed a computational approach to study the effect of neurostimulation on the hippocampus, a brain structure particularly implicated in memory and Alzheimer’s disease. Specifically, they developed a computer model to represent hippocampal neurons in a biologically realistic way, and to simulate the effect of electrical stimulation on their activity (https://elifesciences.org/articles/87356). However, this approach is currently limited because it requires very large computing capacities, and the most widespread software does not allow simulations to be carried out in a reasonable time. At the Material-to-System Integration (IMS) Laboratory, Dr. Levi has recently developed a real-time neuromorphic system for the simulation of biomimetic neurons. The combined expertise of the IMN and IMS are complementary and provide an innovative solution for the simulation and understanding of neuronal mechanisms.
In this context, the RT-HippoNeuroStim project aims at translating the hippocampal model previously developed at the IMN onto the new neuromorphic computing architecture developed at the IMS. This architecture is based on Field Programmable Gate Arrays (FPGA) and shows much faster computation speed than current software emulation. We will leverage this platform to simulate the activity of the hippocampus in real time, which will greatly accelerate research on hippocampal neurostimulation.

Project description: real-time FPGA implementation of a conductance-based hippocampal model

From the previous neural network system implemented on a SoC FPGA, referred to as Bioemus (https://www.biorxiv.org/content/10.1101/2023.09.05.556241v1.abstract), the candidate will adjust the parameters and adapt the BioEmus system to replicate the simulations performed by the Brian2 software. Validation will follow a step-by-step methodology: one network at a time. Subsequently, the development of new processes for synapses and FPGA networks will be undertaken.

Candidate profile

Enthusiastic and curious candidates with a certain autonomy and willing to develop skills at the interface between embedded system and neuroscience are encouraged to apply. The candidate will have knowledge in programming (Python, C/C++), embedded system (C/C++, basic Linux), FPGA design (VHDL, HLS), and computational neuroscience. Hence, multidisciplinary work capability is highly recommended.

Applications

Applications should include a motivation letter, full curriculum vitae, a copy of the relevant diplomas showing marks. If possible, contact details of 2 references.

Contact

• ,
• ,