Lieu : salle George Boole 2 Programme 9h30 : Salli Moustapha. Shared Memory Parallelism for 3D Cartesian Discrete Ordinates Solver 10h30 : Pause café 10h45 : Jérémie Gaidamour. A parallel multigrid framework for implementing grid transfer operators based on constrained minimization 11h45 : Point d’équipe 12h00 : Fin des hostilités …
July 3rd Runtime system – Salle Ada Lovelace Bât. Inria 9h30 10h30 R. Namyst (Inria) – Programming modern (accelerator-based) multicore machines: a runtime system’s perspective 11h00 -12h00 G. Bosilca (Inria & UTK) – Parallel Runtime Scheduling and Execution Controller (ParsEC) July 4th hierachical linear solver – Amphi LaBRI 9h30 – …
Lundi 10 décembre GT à 16h en salle George Boole Abstract : During the last decade, computer architectures for high performance computing have considerably evolved toward heterogeneous systems equipped with different types of computational units and a higher number of cores per chips. An example of popular heterogeneous architectures widely …
GT le 26/11 à 14h15 – salle George Boole 2 This talk will present the last developments made in linear algebra algorithms on top of the PaRSEC runtime, formerly known as DAGuE. 1) The first algorithm presented will be the hierarchical QR. This algorithms exploits the stability of the QR …
GT le 19/09 à 11h – salle George Boole This talk presents a new approach for deriving a robust preconditionner with low memory requirement based on the block LDU factorization. This preconditioner rely on an original representation of the blocks of the L, D and U factors as the sum …
