RED-SEA Project: Towards a new-generation European interconnect

Autor: Gomez, Maria Engracia, Sahuquillo, Julio, Biagioni, Andrea, Chrysos, Nikos, Berton, Damien, Frezza, Ottorino, Lo Cicero, Francesca, Lonardo, Alessandro, Martinelli, Michele, Paolucci, Pier Stanislao, Pastorelli, Elena, Simula, Francesco, Turisini, Matteo, Vicini, Piero, Ammendola, Roberto, Chiarini, Carlotta, De Luca, Chiara, Capuani, Fabrizio, Castelló, Adrián, Duro, Jose, Stabile, Eugenio, Quintana, Enrique, Bernier-Bruna, Pascale, Chen, Claire, Lagadec, Pierre-Axel, Pichon, Gregoire, Walter, Etienne, Katevenis, Manolis, Bartzis, Sokratis, Mousouros, Orestis, Xirouchakis, Pantelis, Mageiropoulos, Vangelis, Gianioudis, Michalis, Loukas, Harisis, Ioannou, Aggelos, Kallimanis, Nikos, de la Rosa, Miguel Sanchez, Gomez-Lopez, Gabriel, Alfaro-Cortés, Francisco, Sahuquillo, Jesus Escudero, Garcia, Pedro Javier, Quiles, Francisco J., Sanchez, Jose L., De Gassowski, Gaetan, Hautreaux, Matthieu, Mathieu, Stephane, Moreau, Gilles, Perache, Marc, Taboada, Hugo, Hoefler, Torsten, Schneider, Timo, Barnaba, Matteo, Brandino, Giuseppe Piero, De Giorgi, Francesco, Poggi, Matteo, Mavroidis, Iakovos, Papaefstathiou, Yannis, Tampouratzis, Nikolaos, Kalisch, Benjamin, Krackhardt, Ulrich, Nuessle, Mondrian, Frings, Wolfang, Gottwald, Dominik, Guimaraes, Felime, Holicki, Max, Marx, Volker, Muller, Yannik, Clauss, Carsten, Falter, Hugo, Huang, Xu, Barillao, Jennifer Lopez, Moschny, Thomas, Pickartz, Simon

Publikováno v: In Microprocessors and Microsystems October 2024 110

Generating and Scaling a Multi-Language Test-Suite for MPI

Autor: Adam, Julien, Besnard, Jean-Baptiste, Canat, Paul, Shende, Sameer, Taboada, Hugo, Roussel, Adrien, Pérache, Marc, Jaeger, Julien

Publikováno v: EuroMPI'23
EuroMPI'23, Sep 2023, Bristol, United Kingdom

International audience; High-Performance Computing (HPC) is currently facing significant challenges. The hardware pressure has become increasingly difficult to manage due to the lack of parallel abstractions in applications. As a result, parallel pro

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=od______3515::a2235036cf727ab9d3c2e88b63e720b9
https://inria.hal.science/hal-04156064

Performance Improvements of Parallel Applications thanks to MPI-4.0 Hints

Autor: Moraru, Maxim, Roussel, Adrien, Taboada, Hugo, Pérache, Marc, Jaillet, Christophe, Krajecki, Michaël

Publikováno v: IEEE SBAC-PAD 2022-34th International Symposium on Computer Architecture and High Performance Computing
IEEE SBAC-PAD 2022-34th International Symposium on Computer Architecture and High Performance Computing, Nov 2022, Bordeaux, France

International audience; HPC systems have experienced significant growth over the past years, with modern machines having hundreds of thousands of nodes. Message Passing Interface (MPI) is the de facto standard for distributed computing on these archi

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::26632b3c5a18224e49df284fca129468
https://doi.org/10.1109/sbac-pad55451.2022.00038

Benefits of MPI Sessions for GPU MPI applications

Autor: Moraru, Maxim, Roussel, Adrien, Pérache, Marc, Taboada, Hugo, Jaillet, Christophe, Krajecki, Michaël

Publikováno v: EuroMPI '21-28th European MPI Users' Group Meeting
EuroMPI '21-28th European MPI Users' Group Meeting, Sep 2021, Leibniz, Germany
EuroMPI '21-28th European MPI Users' Group Meeting, Sep 2021, Leibniz, Germany. ⟨10.1145/1122445.1122456⟩

International audience; Heterogeneous supercomputers are now considered the most valuable solution to reach the Exascale. Nowadays, we can frequentlyobserve that compute nodes are composed of more than one GPUaccelerator. Programming such architectur

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::8f8459b88d936d3616053c11e2a394bb
https://cea.hal.science/cea-03322976

Impact du placement des threads de progression pour les collectives non-bloquantes

Autor: Taboada, Hugo

Publikováno v: Compas 2016: conférence d'informatique en Parallélisme, Architecture et Système
Compas 2016: conférence d'informatique en Parallélisme, Architecture et Système, Jul 2016, Lorient, France

National audience; Les collectives MPI non-bloquantes permettent de recouvrir les communications par du cal-cul. Un bon taux de recouvrement est obtenu en effectuant les tâches de communication et destâches de calcul en parallèle. Pour réaliser c

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::4403900c03c9b329dac4979b1acf6517
https://hal.archives-ouvertes.fr/hal-01355140/file/compas2016.pdf