High Performance Computing in Science and Engineering '09: by Wolfgang E. Nagel PDF

on the finish of the yr 2008, we've seen a strategic step in the direction of a funct- ning HPC infrastructure on Tier-0 point in Germany. in accordance with an contract ( Verwaltungsabkommen ) among the Federal Ministry of schooling and study (BMBF) and the kingdom ministries for examine of Baden-Wurttem- ] berg, Bayern, and Nordrhein-Westfalen, the cheap of total four hundred Million Euro were allotted both shared among federal and kingdom professionals in a ?ve 12 months time-frame to set up the subsequent iteration of HPC platforms on the Gauss Centre for Supercomputing (GCS) along with the 3 nat- nal supercomputing centres HLRS (Stuttgart), NIC/JSC (Julich), and LRZ ] (Munich). As a part of that strategic initiative, in may well 2009 already NIC/JSC has put in the ?rst section of the GCS HPC Tier-0 assets, an IBM Blue Gene/P with approximately three hundred. 000 Cores, this time in Julic ] h, With that, the GCS offers the main powerfulhigh-performance computing infrastructure in - rope already this day. HLRS and its companions within the GCS have agreed on a typical approach for the deploy of the subsequent new release of cutting edge HPC platforms. Over the following couple of years, HLRS and LRZ because the different GCS facilities will improve their platforms consequently. The plan is to have a Tier-0 HPC approach inside GCS working at any time during this ?ve 12 months interval. Asanintermediatestep, HLRShasreplacedmostoftheirNECSX-8nodes by means of the NEC SX-9/12M192, a method with approximately 20 TFLOPs peak."