Cray today announced the availability of the Cray X1E supercomputer, a major upgrade to the company's record-setting Cray X1 scalable vector product. The new system, which features the world's fastest processors and highest computing density, nearly triples the peak performance and price performance of the predecessor Cray X1 system that has set records on weather, engineering and scientific research applications. Initial Cray X1E supercomputer customers will use the systems to handle the most challenging high-performance computing (HPC) problems and workloads at government and research sites in the U.S., Europe and Asia: -- The U.S. Department of Energy's Oak Ridge National Laboratory (ORNL) plans this year to acquire a Cray X1E supercomputer with peak performance of 20 teraflops (trillions of calculations per second), along with a 20-teraflop Cray XT3 massively parallel processing system, as part of the DOE's plan to build the world's most powerful supercomputing capability for open (non-classified) scientific research at ORNL. -- The Korea Meteorological Administration (KMA) will operate one of the largest numerical weather forecasting systems in the world when it upgrades later this year to a Cray X1E supercomputer with peak performance of more than 16 teraflops. As part of a previously announced five-year contract, Cray and KMA have also jointly established the Earth System Research Center to advance atmospheric modeling in the East Asia Pacific region. -- In December 2004, Warsaw University's Interdisciplinary Centre for Mathematical and Computational Modeling (ICM) became the first customer to receive a Cray X1E system. This major upgrade will be used to enhance national weather forecasts for Poland and to support leading-edge research in the mathematical, natural and computational sciences, including bioinformatics. -- Spain's National Institute of Meteorology has ordered a Cray X1E supercomputer as an upgrade to the institute's Cray X1 system. The upgraded system will be used for operational numerical weather prediction and for research in climate and atmospheric modeling. D. I. Lee, KMA's Chief, Supercomputer Center Branch, said, "I would like to express my satisfaction at the selection of the Cray system as our next supercomputer, based on the real performance of KMA's operational applications. The Cray system as a data processing engine will provide the basic infrastructure to understand the Earth System and to make the information and products to support sound decision-making for national policy and management. KMA expects not only good performance from the Cray system, but also good support from Cray as a company." According to ICM Director Marek Niezgodka, "The Cray X1E supercomputer promises to deliver the two great advantages of using smaller numbers of simple, powerful processors within a balanced architecture: better computational capability and a smaller communications barrier. These attributes are very important for accelerating scientific and engineering progress." The Cray X1 predecessor system has turned in record results on large, challenging scientific and engineering applications: -- An Army High Performance Computing Research Center (AHPCRC) unstructured mesh computational fluid dynamics (CFD) application ran at a sustained speed exceeding one teraflop on a Cray X1 supercomputer with 256 multistreaming processors. The AHPCRC demonstrated the use of the Fifth Generation Mesoscale Weather Forecast Model (MM5) on the Cray X1 system to produce 5-kilometer resolution weather forecasts for the entire Continental United States (CONUS), and to produce a 2.5-kilometer CONUS simulation for a 12-hour time period with 15-second time steps (the results of the 2.5-kilometer study are being analyzed to verify the correctness of MM5 at this fine grid spacing). -- Researchers at Oak Ridge National Laboratory have run scientific applications up to 25 times faster than before with the Cray X1 system and achieved a 50 percent performance improvement, on a processor-to-processor basis, over Japan's Earth Simulator on the widely used parallel ocean programme (POP v1.4.3). -- Overall customer-reported scores for the Cray X1 supercomputer system are also the best for any high-performance computing (HPC) system on the new HPC Challenge benchmark tests. "The Cray X1E supercomputer builds on the success of the Cray X1 system by adding binary-compatible, advanced dual-core processors to deliver better performance and greater density--while continuing to scale to thousands of processors in a single system image. It offers a significant improvement in price performance over its predecessor," said Peter Ungaro, Cray senior vice president of sales, marketing and service. About The Cray X1E Supercomputer System The new Cray X1E supercomputer nearly triples the peak performance of its predecessor, boosting the peak processing power available in a single cabinet from 0.819 teraflops to 2.3 teraflops--while maintaining the same cabinet size to produce the highest computing density in the HPC market. To achieve that density, the Cray X1E system employs the world's most powerful individual processors (18.0 gigaflops) and fits two of these into each socket in a dual-core design. The system also boosts memory performance by 50 percent, resulting in five times the processor performance, five times the memory performance and 25 times the interconnect performance of conventional HPC systems. The Cray X1E supercomputer scales to sizes as large as 147 teraflops, and its high-bandwidth, low-latency interconnect enables substantially more efficient scaling than on clustered SMP systems. Cray X1E systems are available with 16 to 8,192 processors and 32 gigabytes to 32 terabytes of memory. Cray offers both air-cooled and liquid-cooled versions. The Cray X1E supercomputer is aimed at the critical computing needs of classified and unclassified government, academic research, and the weather-environmental, automotive, aerospace, chemical and pharmaceutical markets. The system combines the exceptional processor performance of traditional vector systems with the scalability of microprocessor-based architectures.