PRACE A Petaflop Computing Research Infrastructure for European Scientists and Engineers. Thomas Eickermann, October 24, 2008, ICT-BIO Brussels

Transcription

1 PRACE A Petaflop Computing Research Infrastructure for European Scientists and Engineers Thomas Eickermann, October 24, 2008, ICT-BIO Brussels

2 Agenda Introduction to PRACE HPC in Europe why PRACE The ESFRI vision PRACE objectives and status Services (also) for bio-medical sciences Challenges identified in the HET scientific case How to access the future PRACE Infrastructure? How to get involved? 2

3 HPC a Key Technology Supercomputers are the tool for solving most challenging problems through simulations Access to capability computers of leadership class is essential for international competiveness in science and engineering Providing competitive HPC services is a continuous endeavor This has been acknowledged by leading industry nations such as USA and Japan since the 1990 ies And in Europe? 3

4 Europe s current position in HPC Aggregated LINPACK Performance in PetaFlops in November Top 500 Lists 75% of European HPC-power is within PRACE Principal Partner countries 4

5 Is there a need for capability computers in ICT-BIO? Some HPC-related statements picked up yesterday The full problem cannot be solved, we have to reduce/focus [After focusing, ] we have enough computing resources to solve it We get the data for free from the USA A world-class computing service will enable Europe to solve problems today that could otherwise only be solved in 3-5 years cooperate and compete with the USA on equal level Top years 5

6 Computational science infrastructure in Europe The European Roadmap for Research Infrastructures is the first comprehensive definition at the European level Research Infrastructures are one of the crucial pillars of the European Research Area A European HPC service impact foreseen: strategic competitiveness attractiveness for researchers supporting industrial development 6

7 capability The ESFRI Vision for a European HPC service European HPC-facilities at the top of an HPC provisioning pyramid Tier-0: 3-5 European Centres Tier-1: National Centres Tier-2: Regional/University Centres tier-0 tier-1 tier-2 PRACE DEISA Creation of a European HPC ecosystem involving all stakeholders HPC service providers on all tiers Grid Infrastructures Scientific and industrial user communities The European HPC hard- and software industry # of systems 7

8 ESFRI Estimated costs Unlike other European Research Infrastructures: Tier-0 resources have to be renewed every 2-3 years Construction cost Mio. every 2-3 years Annual running cost Mio. A truly European challenge also in terms of funding PRACE The Partnership for Advanced Computing in Europe An Initiative created to implement the ESFRI vision of a European HPC service 8

9 PRACE Project Facts Objectives of the PRACE Project: Prepare the contracts to establish the PRACE permanent Research Infrastructure as a single Legal Entity in 2010 including governance, funding, procurement, and usage strategies. Perform the technical work to prepare operation of the Tier-0 systems in 2009/2010 including deployment and benchmarking of prototypes for Petaflops systems and porting, optimising, peta-scaling of applications Project facts: Partners: 16 Legal Entities from 14 countries Project duration: January 2008 December 2009 Project budget: 20 M, EC funding: 10 M PRACE is funded in part by the EC under the FP7 Capacities programme grant agreement INFSO-RI

10 PRACE Project Consortium GENCI Principal Partners tier 0 tier 1 General Partners New Partners - since May of the PRACE Initiative : 10

11 Project work some highlights Legal form, governance, funding and usage, peer review: options identified analysis and consultation with stakeholders in progress Preparation of procurement and installation of the first Petaflops systems Benchmark suite defined; next step:petascaling of selected applications Prototypes for promising architectures for 2009/2010 selected Links with vendors and technology providers for systems beyond 2010 established Training, education and dissemination Summer school on petascaling in Stockholm Industrial seminar in Amsterdam 11

12 The Scientific Case Created by HPCEUR, an international scientific panel, involving experts from various scientific disciplines in need of HPC ( ) Provides challenges and requirements Weather, Climatology, Earth Science Astrophysics, Elementary particle physics, Plasma physics Material Science, Chemistry, Nanoscience Life Science Engineering HET 12

13 German Scientific Case - Status and Requirements Scientific Field (numbers in TeraFlop/s) Climate and Earth System Research >500 Geophysics >1000 Nanostructure Physics >200 Solid-State Physics >1000 Computational Fluid Dynamics >1000 Astrophysics >500 Elementary Particle Physics and Physics of Hadrons and Nuclei >1000 Materials Science >500 Theoretical Chemistry >300 Soft Matter 3 30 >200 Biophysics and Bioinformatics >1000 Plasma Physics >500 A. Bode, W. Hillebrandt, and Th. Lippert: German Scientific Case for the BMBF, 8/

14 Life Sciences in the Scientific case HPC should support three categories of applications Pure Supercomputer projects, demanding fast interconnects (Capability Computing) Massive projects, that could also run distributed on a Grid (Capacity or High Throughput Computing) Time-critical projects, that could in principle also be run on smaller systems but need a short time-tosolution 14

15 Scientific case: Pure Supercomputer Projects Systems Biology Chromatine dynamics Large scale protein dynamics Protein association and aggregation Supramolecular systems Medicine (e.g. multigenic deseases, secondary effects of drugs) 15

16 Scientific case: Massive Projects Proteome-scale studies Genomics-scale simulations Data mining Scientific case: Time-critical Studies New pathologies Disease evolution Lethal secondary effects 16

17 Scientific case: secondary effects PRACE tier-0 centres will enable networks between Life science and other scientific disciplines that have common methodologies and algorithms Chemistry High energy Physics (common methods:particle-particle interaction, mesoscopic modeling) Engineering and Climate simulations (involves systems biology issues) 17

18 Accessing the future PRACE RI Disclaimer: this is work in progress, not yet confirmed Access Model Based on peer-review: the best systems for the best science Three types of resource allocations Test / evaluation access Project access for a specific project, grant period ~ 1 year Programme access resources managed by a community Free-of-charge Funding Mainly national funding through partner countries European contribution Access model has to respect national interests (ROI) 18

19 How to get involved? If you are a national coordinator of HPC activities and your country is not yet a member: Join the PRACE Initiative! Port your code to the PRACE Prototypes Prototypes will mainly be used project-internally, but Prototypes will also be made available to selected users for testing/porting purposes using a peer-review process Make use of the PRACE training events Winter School in Athens in February 2009, details to be released soon 19

20 How to learn more or get in touch? Visit Attend the PRACE Workshop held in parallel with the ICT 2008 in Lyon on November 26. The workshop will include a networking session at the ICT and an evening session and dinner (registration required) Meet PRACE at the SC 2008 in Austin in November Send an to prace-coordinator@fz-juelich.de 20

21 Opportunities ahead PRACE builds upon the HPC expertise of 14 European countries in HPC service provisioning and on projects like DEISA the expressed support of our national governments, the European Commission and many scientific communities (including EMBL-EBI) an excellent team-spirit grown during the past years of HPCEUR, HET, PRACE and other joint endeavors The time is right to boost competitiveness of European research and economy through HPC create services to fulfill the HPC requirements of the upcoming ESFRI infrastructures create and shape the European HPC ecosystem 21

22 22

23 Members of the HPCEUR panel for Life Science and Biochemistry Prof. Dr. Ulrich Hansmann, NIC - FZJ (u.hansmann@fz-juelich.de) Dr Wolfgang Wenzel, Institut für Nanotechnologie (wenzel@int.fzk.de) Prof. Modesto Orozco UB / BSC (modesto.orozco@bsc.es) Prof. Roderic Guigó UPF (rguigo@imim.es) Prof. Alfonso Valencia, UAM (valencia@cnb.uam.es) Prof. José María Carazo, UAM (carazo@cnb.uam.es) Thomas Simonson, CNRS (thomas.simonson@polytechnique.fr) FR Martin Field, CEA-DSV (martin.field@ibs.fr) Olivier Poch, CNRS (poch@igbmc.u-strasbg.fr) Prof Mark Sansom, University of Oxford (mark@biop.ox.ac.uk) Dr Charles Laughton, University of Nottingham (charles.laughton@nottingham.ac.uk) 23