Annual meeting
Inria Paris, November 7-8, 2016
Salle Jacques-Louis Lions
Program for Monday, November 7
9h00 - Welcome of participants with coffee and brownies
9h30-10h00 - Opening – Stéphane Lanteri
10h00-10h45 - Guest talk
José Maria Cela
Director of CASE (Computer Applications in Science and Engineering) Department
Barcelona Supercomputing Center (BSC)
Exascale simulations
10h45-11h00 - Coffee break
Numerical linear algebra session
11h00-11h30 - Matthieu Kuhn, Hiepacs
An overview of the parallel hybrid solver Maphys: numerics, implementation and applications
11h30-12h00 - Emmanuel Agullo, Hiepacs
A task-based sparse and data-sparse solver stack
Numerical schemes for PDE models session (part1)
12h00-12h30 - Philippe Helluy, Tonus
Parallelization of a LBM solver with Starpu
12h30-13h00 - Sever Hirstoaga, Tonus
Parallelization of the PIC method in Selalib
13h00-14h30 - Lunch
14h30-15h15 - Guest talk
Stéphane Requena
Head of Innovation, GENCI
Update on national and European HPC facilities and opportunities for Inria
Numerical schemes for PDE models session (part2)
15h15-15h45 - Raphaël Léger, Nachos
High order DG schemes for the DEEP-ER system
15h45-16h15 - Coffee break
Numerical schemes for PDE models session (part3)
16h15-16h45 - Michel Kern, Serena
Algebraic and geometric domain decomposition for groundwater flow
Impact on application challenge
16h45-17h30
Nicolas Bouzat, Tonus
Improved gyroaveraging scheme and non-uniform meshing for Gysela
Julien Bigot, CEA et Maison de la Simulation
BOAST and Starpu, two approaches to improve kernels performance in Gysela
17h30 - End of first day
Program for Tuesday, November 8
9h15-10h00 - Guest talk
Paul Gibbon
Head of Computational Science Division
Institute for Advanced Simulation (IAS), Jülich Supercomputing Centre (JSC)
Computational science at the Jülich Supercomputing Centre
Optimization of performances of numerical solvers session (part1)
10h00-10h30 - Olivier Aumage, Storm
Starpu: a heterogeneous task-based runtime system for scientific applications, a platform for experimenting new programming models
10h30-11h00 - Olivier Aumage, Storm
K'Star: a source-to-source C/C++ - Openmp compiler for transparently leveraging task-based runtime systems within scientific applications
11h00-11h30 - Coffee break
Optimization of performances of numerical solvers session (part2)
11h30-12h00 - Jean-François Méhaut ou Brice Videau, Corse
BOAST : bringing optimization through automatic source-to-source transformations
12h00-12h30 - Philippe Virouleau, Corse
Targeting NUMA architectures: compiler and runtime extensions for Openmp
Resilience session
12h30-13h00 - Luc Giraud, Hiepacs
Hard faults and soft errors: possible numerical remedies in linear algebra solvers
13h00-14h30 - Lunch
14h30-15h15 - Guest talk
Laurent Vanel, IBM Systems
Openpower + Opencapi : innovation through open collaboration
Programming models session (part1)
15h15-15h45 - Christian Perez, Avalon
Component model and MSL, a Multi-Stencil Language
15h45-16h15 - Coffee break
Programming models session (part2)
16h15-16h45 - Jérôme Richard, Avalon
Evaluating the component-task model Comet on Gysela
Impact on application challenge
16h45-17h30 - Bernard Vialay, ANDRA
Les enjeux actuels et futurs de la simulation Haute Performance à l’ANDRA
17h30 - End of second day
Final evaluation meeting
Inria Paris, November 9, 2016
Salle Philippe Flajolet
9h00 - Wecome coffee
9h30-10h00 - General presentation of c2s@Exa, Stéphane Lanteri
10h00-10h30 - Focus 1
Emmanuel Agullo, Hiepacs and Olivier Aumage, Storm
Numerical linear algebra solvers on heterogeneous architectures
10h30-11h00 - Focus 2
Philippe Helluy, Tonus and Olivier Aumage, Storm
Discontinuous Galerkin schemes for plasma physics on multicore architectures
11h00-11h30 - Coffee break
11h30-12h00 - Focus 3
Thierry Gautier, Avalon and Olivier Aumage, Storm
Openmp and runtime systems
12h00-12h30 - Focus 4
Luc Giraud, Hiepacs and Stéphane Lanteri, Nachos
High order numerical schemes and scalable linear solvers for frequency-domain electromagnetics
12h30-13h00 - Demo session
Thibaud Kloczko, SED Inria Sophia Antipolis-Méditerranée
Num3sis@c2s
13h00-14h30 - Lunch
14h30-15h00 - Impact on application challenge: numerical fusion
Guillaume Latu, CEA/IRFM
Gysela@Exascale
15h00-15h30 - Impact on application challenge: radioactive waste management
Marc Leconte and Laurent Loth, ANDRA
High performance simulations with Traces
15h30-17h00 - Disciussions
17h00 - End of the meeting
Midterm evaluation meeting
Inria Place d'Italie, January 13, 2015
Salle Algorithme
9h30 - 10h00 - Welcome coffee
10h00 - 10h30 - General presentation, Stéphane Lanteri
10h30 - 10h50 - Pole 1: Numerical linear algebra, Luc Giraud
10h50 - 11h10 - Pole 2: Numerical schemes for PDE models, Stéphane Lanteri
11h10 - 11h35 - CEA/IRFM, Guillaume Latu
High performance tools to model instabilities in tokamak plasmas
11h35 - 12h00 - ANDRA, Laurent Loth
High performance computing for radioactive waste management
12h00 - 13h30 - Lunch
13h30 - 14h10 - Pole 3: Optimization of performances of numerical solvers
* Runtime support activities, Olivier Aumage
* Static and dynamic processing of numerical data sets, François Pellegrini
14h10 - 14h50 - Pole 4: Programming models
* Component models, Christian Perez
* High level parallel programming models, Thierry Gautier
14h50 - 15h30 - Pole 5: Resilience for exascale computing
* Checkpointing strategies, Frédéric Vivien
* Towards resiliency via numerical remedies, Luc Giraud
15h30 - 15h50 - Software development collaborative platform, Thibaud Kloczko
15h50 - 17h00 - Conclusion and discussion
17h00 - End of the meeting
Annual meeting
Inria Bordeaux - Sud-Ouest, July 10-11, 2014
Julien Bigot, Maison de la Simulation
Continuous integration towards more reproducible science the Gysela5D case
Damien Genet, Bacchus project-team
AeroSol solver for CFD problem for modern architecture
Florent Pruvost, Hiepacs project-team
ADT HPC Collective - MORSE
Matias Hastaran, Hiepacs project-team
EZTRACE - ADT EZPerf: easy performance trace analyser
Raphaël Léger, Nachos project-team
Adaptation of a DGTD solver of Maxwell’s equation to the DEEP-ER architecture
Vincent Lanore, Avalon project-team
Towards a reconfigurable HPC component model
Philippe Virouleau, Moais project-team
An OpenMP source to source compiler for Kaapi and StarPU
Nabil Birgle, Pomdapi project-team
A mixed finite element method for deformed cubic meshes
Xavier Lacoste, Hiepacs project-team
PaStiX and MaPHyS overview
Thierry Gautier, Moais project-team
XKaapi : a runtime for highly parallel (OpenMP) application
Fabier Rozar, Maison de la Simulation et CEA/IRFM
Achieving memory scalability in the Gysela code to fit exascale constraints
Philippe Helluy, Tonus project-team
CLAC: an hybrid OpenCL/MPI Discontinuous Galerkin solver for generic conservation laws
Emmanuel Jeannot, Runtime project-team
Process allocation strategies for runtime optimization
Olivier Aumage, Runtime project-team
Recent developments in StarPU
Rachid El Khaoulani, Nachos project-team
Using MaPHyS with the Traces software
Annual meeting
Inria Sophia Antipolis - Méditerranée, December 9-10, 2013
Guillaume Latu, CEA/IRFM
Numerical tools for ITER applications
Guillaume Sylvand, EADS IW and Hiepacs project-team
HPC and numerical simulation at EADS IW
Olivier Aumage, Runtime project-team
Runtimes for heterogeneous systems
Pierre Ramet, Hiepacs project-team
Scalable linear system solvers
Rachid El Khaoulani, Nachos project-team
Parallelization of the TRACES simulation software
Olivier Aumage, Runtime project-team and Thierry Gautier, Moais project-team
ADT K’Star
Matias Hastaran, Hiepacs project-team
ADT EZPerf
Julien Pedron, Hiepacs project-team
ADT MaPHyS@Exa
Florent Pruvost, Hiepacs project-team
ADT MORSE
Radioactive waste management
Kickoff meeting at ANDRA, April 24, 2013
Guillaume Pépin, ANDRA
ANDRA’s use case for the c2s@Exa IPL
Slides
Laurent Loth, ANDRA
The TRACES software
Slides
Marc Leconte, ANDRA
Ongoing high performance computing initiatives at ANDRA
Slides
Luc Giraud, Hiepacs project-team
On hierarchical parallel sparse linear systems
Slides
Cédric Lachat, Bacchus project-team
The PAMPA tool for parallel mesh partitioning and adaptation
Slides
Souhila Sabit, Sage project-team
Numerical models of reactive transport in porous media
Slides
Christian Perez, Avalon project-team
Programming HPC applications with software components
Slides
Nabil Birgle, Pomdapi project-team
A mixed finite element method for deformed cubic meshes
Slides
Edouard Canot, Sage project-team
Activities of the Sage project-team in relation to the c2s@Exa IPL
Slides
Michel Kern, Pomdapi project-team
Activities of the Pomdapi project-team in relation to the c2s@Exa IPL
Slides - part 1
Slides - part 2
Slides - part 3
Michel Kern, Pomdapi project-team
Maison de la Simulation
Slides
Technical meeting on Meshes and their use in large-scale simulations
Institut Henri Poincaré, May 14, 2013
Stéphane Lanteri, Nachos project-team
c2s@Exa – An INRIA Project Lab on high performance computing for computational sciences
Slides
Mariette Yvinec, Geometrica project-team
CGALmesh - A Simplicial Mesh Generator based on the library CGAL
Slides
Clément Jamin, Geometrica project-team
CGALmesh - Going parallel
Slides
François Pellegrini, Bacchus project-team
The Pampa tool for parallel mesh partitioning and adaptation
Slides
Cécile Dobrzynski, Bacchus project-team
MMG3D - Remailleur iso/aniso de maillage en tétraèdres
Slides
Jean Roger, ANDRA
La problématique des maillages pour la simulation numérique des stockages de déchets radioactifs
Slides
Bruno Lévy, Alice project-team
Parallel meshing by enumerating the vertices of Voronoi cells
Slides
Paul-Louis George, Gamma3 project-team
Remarques sur les grands maillages, comment les construire et/ou comment les éviter
Slides
Philippe Helluy, Tonus project-team
Meshes for Tonus
Slides
Nuclear energy production
Kickoff meeting at Inria Sophia Antipolis - Méditerranée, May 17, 2013
Stéphane Lanteri, Nachos project-team
c2s@Exa – An Inria Project Lab on high performance computing for computational sciences
Slides
Virginie Grandgirard/Guillaume Latu, CEA-IRFM
Julien Bigot, Maison de la Simulation
GYSELA: Exascale needs
Slides
Guillaume Latu, CEA-IRFM
JOREK, a parallel code for modelling non linear MHD in Tokamaks
Slides
Boniface Nkonga/Richard Pasquetti, University of Nice - Sophia Antipolis and Castor project-team
HPC and the Castor project-team
Slides
Jacques Blum, University of Nice - Sophia Antipolis and Castor project-team
The Fusion IPL
Slides
Francis Filbet, Khaliffe project-team
Computing issues related to kinetic models for plasma physics
Slides
Pierre Ramet, Hiepacs project-team
From hybrid architectures to hybrid solvers
Slides
Cédric Lachat, Bacchus project-team
The Pampa tool for parallel mesh partitioning and adaptation
Slides
Christian Perez, Avalon project-team
Programming HPC applications with software components
Slides
Thierry Gautier, Moais project-team
Openmp 4.X to programming HPC applications on heterogeneous architectures?
Slides
Technical meeting on sparse linear system solvers
Inria Antenne Parisienne, September 23, 2013
Rachid El Khaoulani, Nachos project-team
Parallelization of the TRACES software
Slides
Pierre Navaro, Tonus project-team
Selalib, une bibliothèque modulaire pour la simulation du modèle gyrocinétique par une méthode semi Lagrangienne
Slides
Laura Grigori, Alpines project-team
Frédéric Nataf, University of Paris 6, J.L. Lions Laboratory and Alpines project-team
Linear solvers developed in Alpines
Slides - Part 1
Slides - Part 2
Emmanuel Agullo, Hiepacs project-team
Luc Giraud, Hiepacs project-team
Pierre Ramet, Hiepacs project-team
Les solveurs linéaires parallèles dans Hie PACS?: aujourd'hui et demain
Slides - Part 1
Slides - Part 2
Slides - Part 3
Jean-Yves L’Excellent, Roma team
MUMPS : un solveur parallèle multifrontal pour la résolution de Ax=b
Slides
Jocelyne Erhel, Sage project-team
Les solveurs linéaires dans l’équipe Sage
Slides
