Thomas Dufaud
Présentation
Current position : Associate Professor (Maître de Conférences) of University of Versailles (UVSQ) since September 2013 I received a Ph.D in Applied Mathematics from Claude Bernard University (Lyon 1) in 2011. I am assistant Professor and affiliated in Maison de la Simulation and LI-PaRAD laboratory where I am a member of the HPC Team. I authored 25 papers in international journals and conferences since 2010. I co-advised 1 Ph.D (2018). My main researches concern numerical linear algebra, parallel solvers and preconditioners, multilevel methods, and their application in the simulation of physical phenomenon. Since 2014 I also study software engineering for HPC and work on YML and multi-SPMD programming models. From 2018 to 2021, I am visiting scientist in the Programming Environment Research team in Riken R-CCS (Kobe, Japan). I am in charge of the first year of the program MCHPS Msc in UVSQ (2014-2017)(2019-) Research interests : Teaching :
Publications
Publications
Modélisation des écoulements hydrothermaux avec ComPASSColloque annuel GDR HydroGEMM 2023 : Journées scientifiques "Hydrogène du sous-sol: étude intégrée de la GEnèse … à la Modélisation Mathématique", Nov 2023, Pau (FRA), France |
|
Machine Learning for Optimal Compression Format Prediction on Multiprocessor Platform2018 International Conference on High Performance Computing & Simulation (HPCS), Jul 2018, Orleans, France. ⟨10.1109/HPCS.2018.00047⟩ |
|
Design of Data Management for Multi SPMD Workflow Programming Model2018 IEEE/ACM 4th International Workshop on Extreme Scale Programming Models and Middleware (ESPM2), Nov 2018, Dallas, United States. pp.9-18 |
|
Understanding the Performances of SMVP on Multiprocessor PlatformThe 2018 International Conference on Parallel and Distributed Processing Techniques & Applications PDPTA'18, Jul 2018, Las Vegas, United States |
|
Understanding the performances of sparse compression formats using data parallel programming modelHPCS 2017 - International Conference on High Performance Computing & Simulation, Jul 2017, Gênes, Italy. pp.667-674, ⟨10.1109/HPCS.2017.103⟩ |
|
Towards an auto-tuning system design for optimal sparse compression format selection with user expertise2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA), Nov 2016, Agadir, France. ⟨10.1109/AICCSA.2016.7945829⟩ |
|
Coupled flow model and simulation in 3D porous-fractured mediaSIAM Conference on Mathematical & Computational Issues in the Geosciences, Jun 2013, Padova, Italy |
|
ARAS2 preconditioning technique for CFD industrial casesDomain Decomposition Methods (DD20), 2013, San Diego, United States |
|
On a numerical method to compute flow in complex 3D geological fractured porous mediaD1 Seminar, INRIA Rennes, 2013, Rennes, France |
|
On-the-fly Singular Value Decomposition for Aitken's Acceleration of the Schwarz Domain Decomposition Method,PARCO11, Sep 2011, Ghent, Belgium. pp.243 - 250 |
|
A two-level preconditioning framework based on a Richardson iterative processDomain Decomposition Methods (DD21), SAGE team, Inria Rennes and LMNO, Caen, Jun 2012, Rennes, France |
|
Numerical Investigations and Parallel Implementation of the ARAS2 Preconditioning TechniquePARALLEL, DISTRIBUTED, GRID AND CLOUD COMPUTING FOR ENGINEERING 2011, Apr 2011, Adjacio, France. Paper 41, ⟨10.4203/ccp.95.41⟩ |
|
|
Multi-SPMD Programming Model with YML and XcalableMPXcalableMP PGAS Programming Language, Springer Singapore, pp.219-243, 2021, ⟨10.1007/978-981-15-7683-6_9⟩ |
|
|
MYX: Runtime correctness analysis for multi-level parallel programming paradigmsHans-Joachim Bungartz; Severin Reiz; Benjamin Uekermann; Philipp Neumann; Wolfgang E. Nagel. Software for Exascale Computing - SPPEXA 2016-2019, 136, Springer International Publishing, pp.545-567, 2020, Lecture Notes in Computational Science and Engineering, 978-3-030-47956-5 (e-book). ⟨10.1007/978-3-030-47956-5_18⟩ |
A two-level preconditioning framework based on a Richardson iterative processJocelyne Erhel and Martin Gander and Laurence Halpern and Géraldine Pichot and Taoufik Sassi and Olof Widlund. Domain Decomposition Methods in Science and Engineering XXI, 98, Springer, 2014, Lecture Notes in Computational Science and Engineering |
|
Acceleration of Convergence for Domain Decomposition MethodsTopping, B.H.V. and Iványi, P. Computational Technology Reviews, Saxe-Coburg Publications, pp.1-24, 2013, ⟨10.4203/ctr.7.1⟩ |
|
Adaptive Aitken-Schwarz Method for Non Separable Operator on Multiprocessor SystemsParallel Computational Fluid Dynamics: Recent Advances and Future Directions, 2010, Parallel Computational Fluid Dynamics: Recent Advances and Future Directions, 978-1-60595-022-8 |
|
|
ARAS: FULLY ALGEBRAIC TWO-LEVEL DOMAIN DECOMPOSITION PRECONDITIONING TECHNIQUE WITH APPROXIMATION ON COARSE INTERFACES2012 |
|
|
Contribution to the development of Aitken Restricted Additive Schwarz preconditioning and application to linear systems arising from automatic differentiation of compressible Navier-Stokes solutions with respect to the simulation’s parametersNumerical Analysis [math.NA]. Université Claude Bernard - Lyon I, 2011. English. ⟨NNT : 2011LYO10230⟩ |