|
> > Workshop 6
Workshop 6 - Wednesday, June 28, from 14h00 to 17h30
Big Data for materials sciences
The NOMAD Laboratory and new data-based schemes for QM-accurate molecular dynamics
Abstract : I will describe the conception, structure and early achievements of the Novel Materials Discovery (NOMAD) Laboratory, a EC-funded Centre of Excellence developing a Materials Encyclopaedia and Big-Data Analytics and advanced graphics tools for materials science and engineering. The issue of data curation, from archiving to conversion with a flexible metadata structure, and the necessity of bespoke accurate parsing for all major software packages, have emerged as key issues in the first year of the CoE’s operation. An exciting aspect of the forthcoming work plan is the development of analytic tools to handle the data contained in the NOMAD Archive, either to identify previously unnoticed correlations between the database materials, or to boost existing modelling algorithms. As an example of the latter case, I will describe how Bayesian inference can be used to boost first-principles molecular dynamics, by means of machine-learning force fields based on appropriate Gaussian process kernel construction. Analysis reveals that the complexity associated with learning the atomic forces of a target physical system can be quantified, and if found so high that the “Big Enough Data” assumption is ruled out, extrapolation can still be avoided by online learning strategies that concentrate QM subroutine calls where/when “chemically novel” configurations are encountered along the system’s trajectory.
|
Biography : Alessandro has over twenty years experience in materials modelling, interdisciplinarily spanning over Physics and Materials Engineering, with a PhD in either discipline. During his career he has been working in various institutions, including the Universities of Keele, Oxford and Cambridge, the Swiss Federal Institute of Technology (EPFL) in Lausanne, and the University of Trieste. In 2010 he became Professor of Physics at King’s College London. He co-directs the Thomas Young London Centre for Theory and Simulation of Materials and director of the UK-JCMaxwell CECAM Node (partners: KCL, Imperial College, University College London, and the Universities of Oxford and Cambridge). His research interests are centred on the development and application of atomistic modelling techniques involving massively parallel computing, dynamical database generation and machine learning. He uses these techniques to investigate materials’ chemo-mechanical properties such as their resistance to fracture and stress corrosion, as well as nanofabrication approaches such as supramolecular self-assembly. In the late 1990’s, he pioneered the “Learn On The Fly” (LOTF) multi-scale modelling scheme whose latest developments include on-the-fly force learning of first-principle forces from dynamically evolving databases using Bayesian techniques. |
- TERATEC Forum is strictly reserved for professionals.
- Participation to exhibition, conferences and workshops is free (subject to seats available)
- On line registration is obligatory to attend exhibition, conferences or the workshops.
- The Vigipirate security plan being raised to its highest level, it is mandatory to register online in advance and come with an identity card order to participate in TERATEC Forum.
- The badge is free of charge and give you access to all events TERATEC Forum.
Teratec 2016 Forum : High Performance Integrated Computational Materials Engineering
- Thermodynamique des phases métastables by Christophe SIGLI, Senior Scientist in Physical Metallurgy, CONSTELLIUM
- Multiscale modeling of materials starting from the atomic scale by Francois WILLAIME, CEA
- PAMSIM project: Massive parallelism in numerical simulation for mechanics by Ionel NISTOR, head of Simulation tools in mechanics, EDF
- Improve Understanding of Material Science using High Performance Computing by Elisabeth MASSONI, directeur du Centre de Mise en Forme des Matériaux, ECOLE DES MINES DE PARIS
- Accélération de calculs de simulations des matériaux sur GPU by François COURTEILLE, Solution Architect, NVIDIA
- La simulation des matériaux en milieu industriels: quelles perspectives?» Round table animated by Gilles ZERAH, CEA
Teratec 2015 Forum : Big data, multiscale and materials
Teratec 2014 Forum : Material simulation: new materials, usage, and durability
- Du matériaux à la structure, quelques méthodes récentes de simulation by Frederic FEYEL, Responsable du pôle Modélisation & Simulation, SAFRAN
- HPC et modélisation multi-échelle des matériaux de structure des centrales nucléaires by Christophe DOMAIN, Engineer, EDF R&D
- Hydrogen in materials studied by High Performance Computing by Matthieu SALANNE, Maître de conférences, UNIVERSITE PIERRE ET MARIE CURI
- Hydrogen in materials studied by High Performance Computing par Gregory GENESTE, CEA
- Decoherence and electron transfers within proteins: are quantum effects relevant for biology ? by Aurélien DE LA LANDE, Chargé de Recherche CNRS, UNIVERSITE PARIS SUD
Teratec 2013 Forum : Materials Simulation
For any other information regarding the workshops, please contact :
Jean-Pascal JEGU
Tel : +33 (0)9 70 65 02 10
jean-pascal.jegu@teratec.fr
Campus TERATEC
2, rue de la Piquetterie
91680 BRUYERES-LE-CHATEL
France
|
|