Description du projet
Exa-scale computing, software and simulation
DEEP proposes to develop a novel, Exascale-enabling supercomputing platform along with the optimisation of a set of grand-challenge codes simulating applications highly relevant for Europe's science, industry and society. The DEEP System will realise a Cluster Booster Architecture that can cope with the limitations purported by Amdahl's Law. It will serve as proof-of-concept for a next-generation 100 Petaflop/s PRACE production system, striving for independent provision of HPC technology, in particular general purpose Exascale performance supercomputers in Europe. The DEEP concept is based on the duality of an advanced multi-core Cluster system with InfiniBand interconnect complemented by a Booster of Intel many-core MIC processors connected through a Terabit EXTOLL network. A novel open source system software stack along with Cluster Booster adapted programming models, libraries, and performance tools will achieve high productivity and will enable unprecedented scalability on millions of cores. The DEEP hardware and software technology is developed in Europe while the new many-core processor is an essential component of international cooperation. The DEEP concept has the potential to improve the power efficiency of HPC systems by an order of magnitude. Its innovative cooling concept will allow approaching power usage effectiveness values very close to 1. Representative HPC codes from Health and Biology, Climatology, Seismic Imaging, Industrial Design, Space Weather, and Superconductivity will be optimised on DEEP and the extrapolation to millions of cores will be demonstrated. The pan-European DEEP consortium has proven competence to meet the project's massive technological and scientific challenges. DEEP aims at disseminating knowledge amongst major European industrial stakeholders and the entire PRACE consortium through its technical advisory group STRATOS and will contribute to the vision of the PROSPECT association for a European HPC technology platform.
Champ scientifique
- natural sciencescomputer and information sciencessoftwaresoftware applicationssystem software
- social scienceseconomics and businesseconomicsproduction economicsproductivity
- natural sciencesearth and related environmental sciencesatmospheric sciencesclimatology
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringcomputer hardwaresupercomputers
- natural sciencesphysical scienceselectromagnetism and electronicssuperconductivity
Appel à propositions
FP7-ICT-2011-7
Voir d’autres projets de cet appel
Régime de financement
CP - Collaborative project (generic)Coordinateur
52428 Julich
Allemagne
Voir sur la carte
Participants (17)
3000 Leuven
Voir sur la carte
1015 Lausanne
Voir sur la carte
2121 LEFKOSIA
Voir sur la carte
Participation terminée
85622 FELDKIRCHEN
Voir sur la carte
52428 Juelich
Voir sur la carte
80539 Munchen
Voir sur la carte
93053 Regensburg
Voir sur la carte
69117 Heidelberg
Voir sur la carte
08034 Barcelona
Voir sur la carte
31057 Toulouse Cedex
Voir sur la carte
91341 Massy
Voir sur la carte
20692 Yokneam
Voir sur la carte
40033 Casalecchio Di Reno Bo
Voir sur la carte
33020 Amaro
Voir sur la carte
81679 Munchen
Voir sur la carte
28020 Madrid
Voir sur la carte
38057 Pergine Valsugana Trento
Voir sur la carte