AREVA, the CEA and its industrial partners from the Research and Technology Computing Center (CCRT, Centre de calcul recherche et technologie) are investing in a new supercomputer to be installed in May 2012 at the CEA's Major Computing Center (TGCC, Très grand centre de calcul) in Bruyères-le-Châtel, in the greater Paris metropolitan area.
The new supercomputer will give AREVA, Astrium, EDF, Ineris, Safran and CEA the highest level of computing resources to support future project development. Digital simulation is a vital tool in research on power generating plant operations, nuclear reactor design and safety, environmental risk assessment, new materials research, and more.
Equipped with the most advanced processors available, the new computer will be supplied by Bull. With a peak of 200 teraflops, its evolutionary architecture was sized to meet the needs of current and future partners. The new acquisition boosts the CCRT’s total computing power to more than 500 teraflops.
Commenting on the acquisition, Martha Heitzman, AREVA’s Senior Executive Vice President of Research and Development said “These supercomputers enable AREVA to perform studies for our customers with more accurate modeling data in a shorter amount of time, supplementing data from our technical centers’ models. With these tools, we are able to demonstrate the safety of our reactors more precisely.”
Some concepts
The CEA’s Major Computing Center (TGCC) has all of the computer resources and services needed to host high-performance supercomputers.
Flops is the acronym of FLoating point Operations Per Second. It is a unit used for measuring the speed or the calculation capacity of a processor. 1 teraflops corresponds to one trillion operations per second. Peak power represents the maximum theoretical power of all of the computer’s processors.