By continuing to browse the site, you are agreeing to our use of cookies. (More details)
Portal > Offres > Offre UPR8001-ANNHEM-004 - H/F Evaluation des méthodologies de type réseau neuronal pour la compréhension de la formation des couches d'interface dans les matériaux intégrés de la microélectronique

H/F Evaluation of neural network methodologies for understanding the formation of interface layers i

This offer is available in the following languages:
Français - Anglais

Ensure that your candidate profile is correct before applying. Your profile information will be added to the details for each application. In order to increase your visibility on our Careers Portal and allow employers to see your candidate profile, you can upload your CV to our CV library in one click!

Faites connaître cette offre !

General information

Reference : UPR8001-ANNHEM-004
Workplace : TOULOUSE
Date of publication : Monday, December 02, 2019
Type of Contract : FTC Technical / Administrative
Contract Period : 18 months
Expected date of employment : 1 May 2020
Proportion of work : Full time
Remuneration : Engineer - Gross monthly between 2000 and 2100 euro
Desired level of education : 5-year university degree
Experience required : Indifferent


The mission is part of a simulation research project in which we want to evaluate the capabilities of the so-called "neural networks" methodologies applied to the case of silicon oxidation modeling.
Many ab initio type structures relevant for the simulation of silicon oxide growth from silicon substrate are available. These structures all of interest are too numerous to be treated and analyzed "humanly". We propose here to evaluate the capacity of "neural network" methods to draw a logical map of these multiple structures.


Task 1 - State of the art and choice
- list the cases where the "neural network" methods have been validated for various applications, particularly in the context of microstructural evolution.
- Choice of the format of the neural network and constitution of the database of the elements useful for the good behavior of the scheduling by the neural network.

Task 2 - Application to the simulation of silicon oxidation
From ab initio data available at LAAS-CNRS, the neural network will first be adapted and then applied.

Task 3 - The potential created on the basis of neural networks can then be used with the kART methodology developed in Montreal to create an Si / SiO2 ab initio interface. It will also be possible to have new atomistic mechanisms to implement in kinetic Monte Carlo models.


Modeling / simulation at the atomic scale of type DFT or DM
Programming, familiar with numerical methods inspired by artificial intelligence (Gaussian methods, neural network, linear regression ...)
Phyisico-chemistry of materials, materials sciences

Work Context

The work will take place at LAAS-CNRS integrated into the M3 team. Simulations will be performed on the CALMIP mesocenter.

Constraints and risks


We talk about it on Twitter!