We always welcome applications for PhD and postdoc positions. Do not hesitate to send us your CV!
Post-doc: Electric-field-gated spin Hall oscillators and diodes for neuromorphic computing
The goal of the study is to fabricate electrical-field gated spin-Hall oscillators and diodes that will mimic non-volatile, tunable neurons and synapses, to characterize and optimize their transport and dynamic properties for neuromorphic computing, and assemble them in hardware neural networks.
More information: https://emploi.cnrs.fr/Offres/CDD/UMR137-JULGRO0-015/Default.aspx?lang=EN
Master 2 Internship: Learning in neural networks made of synchronized nano-oscillators
Small networks of synchronized nano-oscillators perform pattern recognition through their coupled dynamics (Nature 2018). But novel algorithms need to be designed to train large scale networks of coupled oscillators on complex tasks. Equilibrium Propagation is an algorithm very well suited for this purpose, as it finds the synaptic weight updates required for training a physical system directly from its dynamics. The goal of this internship, at the interface between physics and artificial intelliegnce, is to match the two approaches, by simulating a neural network of synchronized nano-oscillators and training it with Equilibrium propagation. The internship can be followed by a Ph.D thesis (simulations + experimental realization of the network) funded by the ERC project Grenadyn.
To apply, send an email to Julie Grollier