Phd in nano/micro-electronics - Artificial neurons based on Mott insulator: realization, testing and modelling

4 Apr 2024
Job Information

Organisation/Company

Nantes Université

Department

IMN UMR-CNRS 6502

Research Field

Engineering » Electronic engineering
Engineering » Materials engineering
Physics » Electronics
Physics » Solid state physics

Researcher Profile

First Stage Researcher (R1)

Country

France

Application Deadline

31 Jul 2024 - 23:59 (Europe/Paris)

Type of Contract

Temporary

Job Status

Full-time

Hours Per Week

35

Offer Starting Date

1 Oct 2024

Is the job funded through the EU Research Framework Programme?

Not funded by an EU programme

Is the Job related to staff position within a Research Infrastructure?

No

Offer Description

Context

For half a century, the information technology revolution has been closely associated with the development of computing. However, all computers are built according to a so-called Von Neumann architecture, where processors and memory are separate blocks. This results in high energy consumption for the processing of massive data. Researchers are working on alternative architectures to develop a more energy-efficient artificial intelligence. The most promising architecture is based on the mammalian brain, which can be 10,000 times more energy efficient than current computers. Currently, intense international competition is growing to realize hardware-based artificial neural networks using artificial synapses and neurons made from innovative materials.

In this context, a team of researchers from IMN has discovered a new property of Mott insulators. This property makes it possible to realize a new type of non-volatile multi-level memories (similar to artificial synapses) and an artificial neuron of the Leaky-Integrate-and-Fire (LIF) type with these materials. The IMN team very recently obtained substantial funding from the Region/CNRS/Nantes University, enabling to finance PhD theses and Post-Docs in order to create the first hardware neural network demonstrator based on Mott insulators. This project aims to lay the foundations for energy-efficient artificial intelligence using these innovative materials.

PhD project

The PhD thesis will focus on an important part of this project. The main objective is to realize, test and model artificial neurons and artificial neural networks based on Mott insulators. The thesis will consist of three parts. 1) Sizing and realization of artificial neurons: the student will define the optimal design of the individual components for their integration into a functional neural network. For this, he might carry out multi-physics simulations, will deposit thin layers by reactive magnetron sputtering and will characterize the produced devices by electrical transport, SEM, TEM, Raman, AFM, XRD… 2) Electrical tests of neurons and neural networks: the student will have to define the protocol for programming artificial neurons, targeting excellent reproducibility of the component characteristics. Within the framework of the overall project, he will also participate in the electrical testing of the neural network fabricated using the optimized components. These tests will be carried out on existing IMN equipment and on a specific probe station acquired as part of the project. 3) Modeling of neurons and the neural network: the student will create a model of the individual neurons behavior which will be integrated into a simulation of the neural network. This will allow him to extrapolate the behavior of the tested network, and to simulate wider / deeper networks.

 

Requirements

Research Field

Engineering » Electronic engineering

Education Level

Master Degree or equivalent

Research Field

Physics » Electronics

Education Level

Master Degree or equivalent

Skills/Qualifications

The candidate will mainly work at the Institut des Matériaux de Nantes Jean Rouxel (the IMN) (https://www.cnrs-imn.fr/ ). It is one of the main materials research centers in France. It now brings together more than 120 researchers (chemists, physicists, materials engineers from the CNRS and Nantes University), administrative and technical staff, and 80 research contractors. Deposition and annealing equipment are available at the IMN, as well as many characterization techniques (XRD, SEM, HR-TEM, XPS, electrical measurements at the single chip level). He/she will work in the “Physics of Materials and Nanostructures” team. The attached doctoral school is the ED3M whose themes cover physics and chemistry in the broad sense.

Languages

ENGLISH

Level

Good

Additional Information
Work Location(s)

Number of offers available

1

Company/Institute

IMN UMR-CNRS 6502

Country

France

City

NANTES

Postal Code

44000

Street

2 Rue de la Houssinière

Geofield

Where to apply

Website

https://theses.doctorat-bretagneloire.fr/3mg/campagne-2024/neurones-artificiels…

Contact

City

NANTES

Website

https://www.cnrs-imn.fr/
http://WWW.UNIV-NANTES.FR

Street

1 QUAI DE TOURVILLE, BP 13522, 44035 NANTES CEDEX 01

E-Mail

[email protected]
[email protected]
[email protected]

STATUS: EXPIRED