Cerfacs Enter the world of high performance ...

Algo-Coop Internal Meetings

The Algo-Coop team meetings are short UNFORMAL seminars of about 20 minutes, dedicated to the introduction of the work done or in progress by the Parallel Algo-Coop team members. They are useful for presenting work you did before arriving at CERFACS, for rehearsing before a conference, for sharing questions on current research with your colleagues…

Beginners are particularly welcome !

We remind you that these meetings are required to be english-spoken.

You can register at this mail address : yzel@cerfacs.fr

                                                                                                                                                                                                          

2021

  • Implementing block preconditioners for multiphysics problems in firedrake, Ana ORDONEZ
  • Fast Linear Solvers for Compatible Discrete Operator Schemes Arising in Incompressible CFD Simulations, Yongseok JANG
  • Latent Space Data Assimilation by using Deep Learning, Mathis PEYRON
  • Background error covariance matrix estimation from multifidelity ensembles, Jérémy BRIANT & Mayeul DESTOUCHES
  • Fast solvers for robust discretizations in computational fluid dynamics, Pierre MATALON (Thesis defense rehearsal)
  • Parallelisable preconditioners for saddle point weak-constraint 4D-Var, Jemima TABEART (University of Edinburgh)
  • Background error covariance matrix estimation from multifidelity ensembles, Jérémy BRIANT
  • Deep learning for species recognition to help protect bears and an embedded network, Elsa GULLAUD
  • Multilevel Monte Carlo (MLMC) methods, Rob Scheichl (U. Heidelberg)
  • Continuity equations and super-resolution microscopy for the reconstruction of a cell membrane potential, Alfio BORZI (U Wuerzburg)
  • Two recent developments in Multilevel Monte Carlo, Mike GILES (University of Oxford)
  • A multigrid-inspired approach for the Augmented Block Cimmino Distributed solver, Philippe LELEUX
  • Algebraic multigrid preconditioner for statically condensed systems arising from lowest order hybrid discretizations,Pierre MATALON
  • A block minimum residual norm subspace solver for sequences of multiple left and right-hand side linear systems, Yanfei XIANG
  • Accounting for hydrometeor variables in a variational ensemble data assimilation scheme applied to the weather prediction model AROME, Mayeul DESTOUCHES
  • Active learning for LSTM neural networks, Jérémy BRIANT

 




 Past seminars 

 

NEWS

ISAF project : To accelerate the transition to SAFs, SAFRAN and CERFACS have together obtained an allocation of 44 million CPU hours on the LUMI-C super-computer as part of the EuroHPC Regular Access call in November 2022.

CERFACS |  11 March 2023

The objective of the ISAF project is to study the impact of SAFs (Sustainable Aviation Fuels) on engine operation and pollutant emissions, using high-fidelity Large Eddy Simulations (LES). The methodology developed at CERFACS combines Analytically Reduced Chemistry (ARC) with a multi-component evaporation model to capture the effect of fuel. The target configurations range from academic burners (CRSB at CORIA, SSB at DLR Stuttgart), to isolated industrial injection systems (MICADO at ONERA, HERON at CORIA) and complete annular combustor configurations (BEARCAT at SAFRAN, NTNU test bench).Read more


A CERFACS article distinguished in Scilight (https://doi.org/10.1063/10.0017474)

CERFACS |  9 March 2023

The recently published paper in Physics of Plasma: "3D particle-in-cell study of the electron drift instability in a Hall Thruster using unstructured grids," by W. Villafana, B. Cuenot, and O. Vermorel ( attracted the attention of Scilight (), whose goal is to present the most interesting research in the physical sciences published in AIP journals."Read more

ALL NEWS