The Maison de la Simulation is a research laboratory about numerical simulation. Those activites concern transverse fields like IT and visualization, as well as HPC(High Performance Computing.

The Maison de la Simulation is involved in projects for research and supporting scientific community purpose : tools development, ANR, European projects ...

Information on all HPC training proposed by Maison de la Simulation. These trainings are organized in the context of the French PATC (PRACE Advanced Training Center), with collaborators like IDRIS, Equip@meso or Groupe Calcul

With the development of High Performance Computing infrastructure, both at the national (GENCI in France) and European (PRACE) level, researchers now have access to world-class supercomputers. Such supercomputers should not only make possible major breakthroughs in various scientific domains, but also lead to important breakthroughs impacting industrial competitiveness. Achieving these aims requires using tools designed and optimized for a very high level of parallelism. Moreover, exploiting the large amount of data resulting from large simulations requires appropriate post-processing and visualization tools.

The complexity of Petaflop/s system architectures and those of the targeted simulations represent a real challenge. Tackling this challenge requires building interdisciplinary teams including HPC (High Performance Computing) computer scientists, numerical analysts and experts in the targeted application fields.


This diagnosis led the CEA, the CNRS and the INRIA to create the Maison de la Simulation whose objective is to promote the efficient use of the large HPC equipments by the scientific community, as well as research in the HPC field.


The Maison de la Simulation is now a joint project of five partners (CEA, CNRS, INRIA, University of Orsay and University of Versailles-StQuentin) with the status of a "Unité de Service et de Recherche" (Service and Research Unit), whose aim is to support and stimulate the scientific community in order to get the best out of supercomputers, in particular those managed by the French GENCI and the European PRACE programs. The Maison de la Simulation promotes among others the emergence in France of a HPC community, and develops the strong synergies between researchers and engineers from various fields, necessary for the important scientific breakthroughs expected from HPC to materialize. These initiatives are targeted to the current HPC users, as well as to the research of new application fields for the HPC.



PhD: Design of a top-down computer vision algorithm dedicated to the recognition of cortical sulci
We are seven billion humans with unique cortical folding patterns. The cortical folding process occurs during the last trimester of pregnancy, during the emergence of cortical architecture. The folding patterns are impacted by architectural features specific to each individual. Hence, they could reveal signatures of abnormal developments that can lead to psychiatric syndroms. For the last 25 years, the image analysis lab of Neurospin has been designing dedicated computer vision tools to tackle More »
PhD: Development of "all-regime" AMR simulation methods for fluid dynamics, application in astrophysics and two-phase flows
Although classic simulation methods for compressible flow are efficient for shock capturing, they are not adapted to variable Mach regimes. Innovative methods using Finite Volume numerical schemes, robust and uniformly accurate with respect to the Mach number (so- called « all-regime »), were recently developed at CEA. These methods allow to solve the equations of compressible flows for both shocks capturing and flows involving very low material speed. Using the ground of these More »
PhD: Advanced scientific vizualisation of massive molecular data on a vizualization wall
Nanosciences as well as life sciences evolve rapidly towards the molecular-level study of objects that are more and more complex. A bottleneck exists in the visual representation of such complexity, that we propose to tackle in the present project. Our goal –in which the PhD to be recruited will play a major role—will be to develop a unique informatics tool for the manipulation and visualization of massive data sets of molecular structures and of their properties which will be made More »
PhD: Optimization of massive data-fluxes in parallel codes in prevision for Exascale, application to the Gysela code
The Ph.D. proposed here aims to study and solve problems faced in handling the massive data fluxes appearing in parallel codes and especially the Gysela code. The goal is to design, implement and evaluate new approaches to tackle seemingly antagonistic aims : (a) optimize performances of data-fluxes, (b) ensure the portability of those performances and (c) improve the readability and maintainability of the code describing them. The first aspect that will be tackled is the handling of More »
PhD: New Particle-in-Cell algorithms for realistic 3D simulations of high-intensity laser-plasma interaction
The success of the PetaWatt (PW) APOLLON laser facility presently under construction, which aim at producing promising particle and light sources from relativistic laser-plasma interactions, will partly rely on the strong coupling between experiments and large-scale simulations with the Particle-In-Cell (PIC) code SMILEI currently under development at CEA/Maison de la Simulation. Standard 3D PIC codes currently in use fail to accurately describe these new interaction regimes because the finite More »