Business Continues to Grow for the Cray XT5m System and the Recently Announced Cray XT6m System

“The Cray XT5m and Cray XT6m systems are industry leaders in terms of compute density and energy efficiency for x86 systems, and our customers – both new and existing – are finding that price and performance efficiency are compelling fits for their needs,” said Barry Bolding, Cray’s vice president of scalable systems.

The Institute of Statistical Mathematics is the first Japanese customer for the Cray XT6m system, which was announced in November 2009 and is expected to be available in the second quarter of 2010.  The Tokyo-based research organization will leverage the Cray XT6m system’s high-density compute power to perform statistical data analysis, statistical modeling, mathematical analysis and statistical interference.

In Europe, the University of Duisburg-Essen is the first Cray XT6m customer in Germany. The University’s researchers and scientists will use the scalable and upgradeable supercomputing system to support its scientific work in chemistry, physics, mathematics and engineering. Research will include the development of parallel algorithms, large-scale computations of electronic structure and molecular dynamics of nano materials, as well as structural mechanics and biomechanics studies. 

“This is the first supercomputing purchase for the researchers and scientists who work on our campus, and we are very excited to now have a system with such impressive abilities,” said Prof. Dr. Eckhart Spohr from the University of Duisburg-Essen. “The simulations that will be performed on our new Cray supercomputer will contribute significantly to the atomistic understanding of structure and reactivity in the nano sciences, energy technology and material sciences, as well as to insight into structural and biomechanical problems.”

Sales of the Cray XT5m system remain strong, with recent wins at multiple weather and climate datacenters, such as the Finnish Meteorological Institute, and at a Fortune 500 pharmaceutical company based in the U.S.

“What truly excites us about these midrange systems is that we now have the ability to provide our production petascale technologies to new market segments that we have not reached before,” said Bolding. “Researchers can now take their science to new levels with a line of proven supercomputers that are the most reliable and upgradeable HPC systems available. With our recent wins across new markets, we are now providing supercomputing resources to a broader base of researchers, scientists and engineers tasked with improving the world around us.”

Introduced in March 2009, the Cray XT5m system builds on the success of the multi-petaflop Cray XT5 supercomputer, whose installations include “Jaguar,” which is the world’s fastest supercomputer and is located at the Oak Ridge National Laboratory. The Cray XT5m supercomputer, the midrange variant of the Cray XT5, is a massively parallel processing system that delivers performance, efficiency, reliability and manageability unrivaled in its price range with capabilities that previously were only available to the world’s largest research facilities.

The Cray XT6m is the company’s second generation of its midrange supercomputer designed to effectively scale down Cray’s high-end systems while providing the same benefits to an expanded base of users. Upgradeable from a Cray XT5m, the Cray XT6m includes compute blades that feature four compute nodes designed for high scalability in a small footprint and can be configured with up to 96 dual-socket nodes per cabinet. Each compute node is composed of two AMD Opteron 6100 Series processors (the eight and 12-core “Maranello” platform), each coupled with its own memory and dedicated Cray Seastar2+ interconnect. The compute nodes in the Cray XT6m systems can also be configured with 32 GB or 64 GB DDR3 memory.

Both of Cray’s midrange systems, with prices starting at $500,000, feature the option of using the company’s industry-leading ECOphlex liquid cooling technology, designed to reduce the customer’s energy usage and lower the total cost of ownership.