Integrated circuit for neuromorphic architecture at CEA LETI (Grenoble)

The candidate should hold a Master in Electrical Engineering, and have some experience in analog integrated circuits design. Knowledge of basic neuroscience would be appreciated, but is not mandatory.

This PhD project is part of Arch²Neu project, funded by the French National Research Agency. This project aims at designing a versatile neuromorphic computing machine. To achieve this, a programmable neuromorphic architecture as well as the associated electronic circuit will be developed. A library of functions that can be executed by this architecture, and a compilation toolchain allowing simple programming of the system will also be developed. From the user’s point of view, this machine should be as simple to program as a classical microprocessor. What is interesting is the fact that this machine would be very well suited to some classes of functions, such as signal processing functions.

The candidate will be in charge of the definition, implementation, and tests of the integrated circuit supporting the architecure. A major part of this project will aim at proposing, given the latest technological advances, an optimal structure for the electronic neuron which behavior is by nature analog. Particularly, we will aim at endowing the electronic neuron with a large number of programmable parameters to extend the class of functions that the system can perform.

The candidate will strongly interact with the other members of the project-team working on architecture design and digital implementation. The candidate will benefit of a 100+ reseachers, leading laboratory in electronics and circuit design.

Informations :

ASIC Design Department
Minatec
CEA-LETI, Grenoble
Desired start date : December 1st, 2009

Please contact
Rodolphe Héliot
rodolphe.heliot@gmail.com
Tel: + 33 6 73 27 75 40

or
Olivier Temam
INRIA Saclay
olivier.temam@inria.fr