{"count": 12, "next": null, "previous": null, "results": [{"tra_keywords": ["HPC", "GPU", "High Performance Computing", "Acc\u00e9l\u00e9rateur"], "tra_speakers": ["Nathanael Schaeffer", "Vincent Boulos"], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "", "tra_id": 120, "tra_title": "Retour d'exp\u00e9rience sur les X\u00e9on-phi\" collaboration BULL", "tra_objective": "", "tra_web": "", "tra_place": "LJK - Grenoble", "tra_start_date": "2013-10-23", "tra_end_date": "2013-10-23", "tra_redirect": false}, {"tra_keywords": [], "tra_speakers": [], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["MaiMoSiNE"], "tra_snippet": "Ce s\u00e9minaire en entr\u00e9e libre est organis\u00e9 par l'Institut Fourier, le LJK, AMIES et MAIMOSINE.\r\n", "tra_id": 95, "tra_title": "Introduction of Mathematics for Industry (Masato Wakayama, Kyushu University)", "tra_objective": "Ce s\u00e9minaire en entr\u00e9e libre est organis\u00e9 par l'Institut Fourier, le LJK, AMIES et MAIMOSINE.\r\n", "tra_web": "", "tra_place": "Institut Fourier, Salle 04", "tra_start_date": "2013-02-22", "tra_end_date": "2013-02-22", "tra_redirect": false}, {"tra_keywords": ["HPC", "M\u00e9thodes Num\u00e9riques", "Calcul Scientifique"], "tra_speakers": [], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "The PETSc (www.mcs.anl.gov/petsc) library offers tools for the parallel solution of partial differential equations, with", "tra_id": 93, "tra_title": "Introduction to PETSc: a user's point of view", "tra_objective": "The PETSc (www.mcs.anl.gov/petsc) library offers tools for the parallel solution of partial differential equations, with an emphasis on scalability.
\r\nPETSc mainly supports MPI parallelism, and is employed in an impressive list of modeling softwares in many different application areas. Its features include parallel vector and matrices, as well as data and grid management tools.
\r\nFor the solution itself, parallel preconditioners and Krylov iterative solvers are offered, as well as interfaces with external packages like the direct solvers PaStiX, MUMPS and SuperLU.
\r\nMoreover, parallel Newton-based nonlinear solvers and timestepping ODE solvers are provided.
\r\n
\r\nThis talk will survey the PETSc main features, and focus on my personal user experience.
\r\n
\r\nIt will also serve as an introduction to the PETSc training planned for the Spring 2013.", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2013-01-22", "tra_end_date": "2013-01-22", "tra_redirect": false}, {"tra_keywords": ["Parall\u00e9lisation hybride", "HPC", "High Performance Computing"], "tra_speakers": ["Philippe Wautelet"], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "L'evolution des supercalculateurs et la multiplication du nombre de coeurs par \r\nnoeud pose un nouveau d\u00e9fi quant a l'ut", "tra_id": 82, "tra_title": "Petite introduction a la programmation hybride MPI/OpenMP", "tra_objective": "L'evolution des supercalculateurs et la multiplication du nombre de coeurs par \r\nnoeud pose un nouveau d\u00e9fi quant a l'utilisation optimale de ces ressources. La \r\nprogrammation hybride MPI/OpenMP qui consiste a m\u00e9langer deux paradigmes de \r\nparall\u00e9lisme est actuellement l'approche privil\u00e9gi\u00e9e pour exploiter ces \r\nnouvelles architectures. Le but de ce s\u00e9minaire est d'introduire cette \r\napproche avec ses avantages et inconv\u00e9nients.
\r\n
\r\nCe s\u00e9minaire se prolongera par une discussion ouverte sur vos probl\u00e9matiques.", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2012-11-27", "tra_end_date": "2012-11-27", "tra_redirect": false}, {"tra_keywords": ["I/O", "Parall\u00e9lisation hybride", "Parall\u00e9lisation", "Parall\u00e8le"], "tra_speakers": ["Yann Meurdesoif"], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": " - Parall\u00e9lisation des I/O, d\u00e9veloppement en cours sur XIOS
\r\n
\r\n - Pr\u00e9sentation d'exp\u00e9riences de parall\u00e9lisation", "tra_id": 64, "tra_title": "Parall\u00e9lisation des I/O, d\u00e9veloppement en cours sur XIOS", "tra_objective": " - Parall\u00e9lisation des I/O, d\u00e9veloppement en cours sur XIOS
\r\n
\r\n - Pr\u00e9sentation d'exp\u00e9riences de parall\u00e9lisation mixte MPI/OpenMP sur d'autres mod\u00e8les (LMDZ, ORCHIDEE, INCA, REPROBUS...) et \u00e9changes avec Yann Meurdesoif.
\r\n", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2012-06-29", "tra_end_date": "2012-06-29", "tra_redirect": false}, {"tra_keywords": ["Outils de D\u00e9veloppement", "D\u00e9bogueur", "Parall\u00e8le", "Software"], "tra_speakers": [], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "Pr\u00e9senter les possibilit\u00e9s et les avantages de l'outil de debugging de code parall\u00e8le Allinea DDT pour le debug \u00ab \u00e0 l\u2019\u00e9c", "tra_id": 60, "tra_title": "Allinea DDT - Le debugging \u00ab \u00e0 l\u2019\u00e9chelle \u00bb de codes parall\u00e8les scalables ", "tra_objective": "Pr\u00e9senter les possibilit\u00e9s et les avantages de l'outil de debugging de code parall\u00e8le Allinea DDT pour le debug \u00ab \u00e0 l\u2019\u00e9chelle \u00bb en l'appliquant \u00e0 des exemples techniques simples et concrets.", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2012-05-30", "tra_end_date": "2012-05-30", "tra_redirect": false}, {"tra_keywords": ["Mod\u00e9lisation Oc\u00e9anique", "Analyse de performances - HPC"], "tra_speakers": [], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "L'\u00e9quipe MEOM: Mod\u00e9lisation des Ecoulements Oceaniques Multi\u00e9chelle (LEGI-CNRS) r\u00e9alise des simulations num\u00e9riques (code", "tra_id": 56, "tra_title": "Calcul massivement parall\u00e8le dans un contexte de mod\u00e9lisation oc\u00e9anique: retour d'exp\u00e9rience", "tra_objective": "L'\u00e9quipe MEOM: Mod\u00e9lisation des Ecoulements Oceaniques Multi\u00e9chelle (LEGI-CNRS) r\u00e9alise des simulations num\u00e9riques (code NEMO) afin de mod\u00e9liser la circulation oc\u00e9anique. Ces simulations sont r\u00e9alis\u00e9es dans diff\u00e9rents centres de calcul.
\r\n
\r\nActuellement la configuration la plus \"lourde\" (CPU-time consuming) est le mod\u00e8le ORCA au 12\u00e8me de degr\u00e9 (1 milliard de points de grille). Dans cette configuration, les simulations sont r\u00e9alis\u00e9es sur le supercalculateur SGI-Altix JADE (GENCI-CINES, Montpellier) sur 2000 \u00e0 3000 coeurs de type Nehalem.\r\n\r\n", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2012-03-28", "tra_end_date": "2012-03-28", "tra_redirect": false}, {"tra_keywords": ["HPC", "GPU", "High Performance Computing"], "tra_speakers": ["Luigi Genovese"], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "In this presentation, we will revise the main aspects of evolution of parallel architectures.", "tra_id": 42, "tra_title": "Parallel Codes and High Performance Computing: General introduction to massively parallelism", "tra_objective": "In this presentation, we will revise the main aspects of evolution of parallel architectures.", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2012-03-14", "tra_end_date": "2012-03-14", "tra_redirect": false}, {"tra_keywords": ["HPC", "Intergiciel", "High Performance Computing", "D\u00e9ploiement sur grille"], "tra_speakers": ["Eddy Caron"], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "L'objectif de cette pr\u00e9sentation sera de montrer comment nous avons\r\nint\u00e9gr\u00e9 l'application de mod\u00e8le atmosph\u00e9rique et qu", "tra_id": 44, "tra_title": "Mod\u00e9lisation atmosph\u00e9rique dans un environnement de grille de calcul : Retour d'exp\u00e9rience. ", "tra_objective": "L'objectif de cette pr\u00e9sentation sera de montrer comment nous avons\r\nint\u00e9gr\u00e9 l'application de mod\u00e8le atmosph\u00e9rique et quelles sont les\r\ntechnologies mises en \u0153uvre pour r\u00e9pondre \u00e0 l'infrastructure de grille\r\ncibl\u00e9e. Dans le cas qui nous int\u00e9resse ici, l'infrastructure cible est CIMENT\r\nle projet de calcul de taille m\u00e9soscopique au sein des Universit\u00e9s\r\nGrenobloise. ", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2012-03-07", "tra_end_date": "2012-03-07", "tra_redirect": false}, {"tra_keywords": ["Biblioth\u00e8que scientifique", "Calcul Scientifique", "Language de programmation", "Langage de programmation", "Fortran", "C++"], "tra_speakers": ["Christophe Picard", "Laurence Viry"], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "", "tra_id": 31, "tra_title": "Fortran2003 / C++avanc\u00e9", "tra_objective": "", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2012-01-25", "tra_end_date": "2012-01-25", "tra_redirect": false}, {"tra_keywords": ["HPC", "Calcul Scientifique", "Language de programmation"], "tra_speakers": ["Christoph Groth"], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["CIMENT", "Groupe calcul Grenoblois", "MaiMoSiNE"], "tra_snippet": "How to choose the right programming language for a specific project involving numerical computation.", "tra_id": 30, "tra_title": "Introduction to programming languages and paradigms", "tra_objective": "How to choose the right programming language for a specific project involving numerical computation.", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2012-01-11", "tra_end_date": "2012-01-11", "tra_redirect": false}, {"tra_keywords": ["Compressible", "Navier-Stokes", "Mod\u00e9lisation"], "tra_speakers": ["Didier Bresch"], "tra_type": "S\u00e9minaire", "tra_type_slug": "seminaire", "tra_frames": ["MaiMoSiNE"], "tra_snippet": "

Dans ce cours, nous d\u00e9crirons les m\u00e9thodes actuelles permettant de montrer le caract\u00e8re bien pos\u00e9 des \u00e9quations de Na", "tra_id": 1, "tra_title": " Sur le caract\u00e8re bien pos\u00e9 des \u00e9quations de Navier-Stokes", "tra_objective": "

Dans ce cours, nous d\u00e9crirons les m\u00e9thodes actuelles permettant de montrer le caract\u00e8re bien pos\u00e9 des \u00e9quations de Navier-Stokes compressible avec r\u00e9gularit\u00e9 faible sur les donn\u00e9es.

\r\n", "tra_web": "", "tra_place": "Grenoble", "tra_start_date": "2011-02-15", "tra_end_date": "2011-02-15", "tra_redirect": false}]}