About PRACE
Building a world-class pan-European High Performance Computing (HPC) Service is a highly ambitious undertaking that involves governments, funding agencies, centres capable to host and manage the supercomputers, and the scientific and industrial user communities with leading edge applications. In contrast to Research Infrastructures that focus on a single scientific instrument an HPC Infrastructure has two unique characteristics: supercomputers serve all scientific disciplines and tier-0 supercomputers have a three year depreciation cycle as tier-0 implies leading edge services.
This requires a periodic renewal of the systems and a continuous upgrade of the infrastructure. Furthermore, novel architectures and system designs will be created by the vendors for leadership systems. There will be several different systems each of them serving a particular application spectrum best.
This fact mandates a distributed Research Infrastructure (RI), since no single site can host all the necessary systems because of floor space, power, and cooling demands. PRACE, the Partnership for Advanced Computing in Europe, is creating a persistent pan-European High Performance Computing RI and related services. Four nations (France, Germany, Italy and Spain) have agreed to provide 400 million Euro to implement supercomputers with a combined computing power in the multi Petaflop/s range over the next 5 years. This funding is complemented by up to 70 million Euros from the European Commission which is supporting the preparation and implementation of this infrastructure. These leadership class systems will help European scientists and engineers to remain internationally competitive.
PRACE will maintain a pan-European HPC service consisting of up to six top of the line leadership systems (Tier-0) well integrated into the European HPC ecosystem. Each system will provide computing power of several Petaflop/s (one quadrillion operations per second) in midterm. On the longer term (2019) Exaflop/s (one quintillion) computing power will be targeted by PRACE. Users will be supported by experts in porting, scaling, and optimizing applications to novel, highly parallel computer architectures. An in-depth training program accompanies the PRACE offering teaching scientists and students how to best exploit the unprecedented capabilities of the systems. A scientific steering committee will provide advice to PRACE and operate alongside a bespoke peer review process through which access to the Tier-0 resources will be granted based on scientific excellence.
The PRACE RI is an international non-profit association with its seat in Brussels. The association is named ‘Partnership for Advanced Computing in Europe AISBL’ and has 19 members, representing Austria, Bulgaria, Cyprus, Czech Republic, Finland, France, Germany, Greece, Ireland, Italy, The Netherlands, Poland, Portugal, Serbia, Spain, Sweden, Switzerland, Turkey and the UK. A Partner from Norway will accede to PRACE by June 30. Additional European states are invited to join.
PRACE RI has already started to operate. The first production system, a one Petaflop/s IBM Blue Gene/P (Jugene) is installed at Forschungszentrum Jülich, Germany, a Gauss Centre for Supercomputing member site. A call for proposals for access to resources on this first PRACE Tier-0 system was opened on May 10, 2010. Jugene is one of the most powerful computers in the world.
More information about PRACE calls is available on www.prace-project.eu/hpc-access
Several other world-class systems will follow the first production system. PRACE has to date secured 400 million Euros of national funding for Tier-0 systems for the next five years. The implementation will be funded in part through European Commission funding in the First Implementation Phase project (PRACE-1IP) starting on July 1, 2010.
A European model of a sustainable high performance ecosystem consists of a small number of supercomputer centres offering computing service at the highest performance level; national and regional centres with supercomputers offering a the performance to run most of the advanced computing; and the local computing centres in universities, research labs or in other organizations strengthening software development and researchers’ competence in computational science.