PRACE Resources

The PRACE RI provides access to distributed persistent pan-European world class HPC computing and data management resources and services. Expertise in efficient use of the resources is available through participating centers throughout Europe.
Available resources are announced for each Call for Proposals.

PRACE Production systems (in alphabetical order of the systems’ names):


GIF - 41.4 kb

CURIE opened for PRACE researchers on 1st March 1st, 2012

The Research Infrastructure PRACE (Partnership for Advanced Computing in Europe) welcomed the full installation of the French Tier-0 supercomputer system CURIE for PRACE, which was completely opened to scientists on 1st of March, 2012. CURIE is a new supercomputer of GENCI, located in France at the Très Grand Centre de Calcul (TGCC) operated by CEA near Paris. It was installed in two phases, starting at the end of 2010 and is now fully operational. During this period, CURIE has gradually been made accessible for research purposes through the PRACE Regular and Preparatory Access Calls. CURIE is a BULL x86 system based on a modular and balanced architecture of thin (5040 blades,each with 2 sockets based on the latest Intel SandyBridge processor), large (90 servers, each with 128 cores and 512 GB of memory) and hybrid (144 blades, each with 288 nVIDIA M2090 GPUs) nodes with more than 360 TB of distributed memory and 15 PB of shared disk. Altogether, CURIE delivers a peak performance of 2 Petaflop/s (2 million billion operations a second).1 “This capacity will help Europe to lead the world in the quest for suitable solutions to societal challenges such as population ageing, climate change and energy efficiency”, said Neelie Kroes, Vice-president of the European Commission for Digital Agenda. It is crucial for PRACE to continuously expand its facilities and services in order to provide the best resources for researchers who use computing science to solve Grand Challenges that can help improve our society and our life on this Earth. Through CURIE the PRACE Research infrastructure has become even better with this new, unique and highly efficient computing architecture.

For technical assistance:


FERMI Photo 10-10-12 15 40 22-300
Italian supercomputer systems complement the PRACE infrastructure from spring 2012. FERMI, the new supercomputing system in CINECA, Italy complements the PRACE Tier-0 supercomputer systems infrastructure. This new system, based on IBM BG/Q architecture, was installed in spring 2012 and became fully operational in August. FERMI was already available for regular calls for awardees and also for subsequent PRACE calls for access. This impressively large Italian supercomputer will be an effective reply to the growing demand from national and European academia, research institutions and industries for Peta-scale computing capabilities. “It will strengthen the position of the PRACE consortium and reinforce Europe’s competitiveness in the race to exa-scale”, said Marco Lanzarini, Director of CINECA. For technical assistance:

Hazel Hen

Hazel Hen

HLRS’s new Hazel Hen-system is powered by the latest Intel Xeon processor technologies and the CRAY Aries Interconnect technology leveraging the Dragonfly network topology. The installation encompasses 41 system cabinets hosting 7,712 compute notes with a total of 185,088 Intel Haswell E5-2680 v3 compute cores. Hazel Hen features 965 Terabyte of Main Memory and a total of 11 Petabyte of storage capacity spread over 32 additional cabinets hosting more than 8,300 disk drives which can be accessed with an input-/output rate of more than 350 Gigabyte per second. For technical assistance:



JPEG - 51.4 kb

Since 1 November 2012, the Gauss Center for Supercomputing provides access to an IBM Blue Gene/Q system JUQUEEN at Forschungszentrum Jülich (FZJ) in Jülich, Germany. Systems of this type are currently the most energy-efficient supercomputers according to the Green 500 List. JUQUEEN has an overall peak performance of 5.87 Petaflop. It consists of 28 racks; each rack comprises 1024 nodes (16394 processing cores). The main memory amounts to 458 TB. More information is available at JUQUEEN’s home page (

For technical assistance:


JPEG - 844 kb

IBM System X iDataplex – MareNostrum – hosted by BSC in Barcelona, Spain.

MareNostrum is based on Intel Sandy Bridge EP processors 2,6 GHz (eight–core), 2 GB/core (32 GB/node) and around 500 GB of local disk acting as local /tmp. All compute nodes are interconnected through an Infiniband FDR10 network,with a non-blocking fat tree network topology and a peak performance of 1 Petaflop/s.

MareNostrum is also providing a rack with Intel MIC accelerators. The configuration per MIC node is:

  • 2 Intel Xeon CPU E5-2670 @ 2.60GHz processors (8 cores/processor)
  • 64 GB of RAM memory
  • 2 Intel Xeon Phi 5110P

For technical assistance:


JPEG - 68.7 kb

SuperMUC is the Tier-0 supercomputer at the Leibniz Supercomputing Centre (LRZ) in Garching, Germany. It provides resource to PRACE via the German Gauss Centre.

The system is an IBM System x iDataPlex based on Intel Sandy Bride EP processors. SuperMUC has a peak performance of 3.2 PFLOP/s consisting of 18 islands, each combining 512 compute nodes with 16 physical cores and 32 GB per node. The nodes are connected by a non-blocking fat tree, based on Infiniband FDR10. Additionally, an island with 205 nodes and 256 GB per node is available. With its innovative warm water cooling system, SuperMUC is one of the most energy efficient supercomputers in the world. For parallel I/O, SuperMUC provides 10 PByte of storage with the GPFS file system from IBM. The batch queueing system LoadLeveler provides novel energy-tags to adjust the CPU clock speed to optimize the “energy to solution”.

For technical assistance: or

Share: Share on LinkedInTweet about this on TwitterShare on FacebookShare on Google+Email this to someone