Cray XD1 supercomputers are designed to take advantage of Dual-Core AMD Opteron processors. With two processor cores on a single die, the dual-core devices boost performance without requiring as much power, cooling or space as two single-core processors. In addition, Rice University will equip its Cray XD1 supercomputer with field-programmable gate arrays (FPGAs) to allow their researchers to explore possibilities for accelerating applications using reconfigurable computing techniques.

"Rice's research community is growing rapidly, and the users have demanded a system that will provide the processing power required to support a diverse set of compute-intensive challenges across both science and engineering”, stated Jan Odegard, executive director of Rice's Computer and Information Technology Institute (CITI). "After performing extensive application benchmarks, we chose the Cray XD1 supercomputer with Dual-Core AMD Opteron processors because it significantly outperformed competing systems when running several key applications.”

"The Cray XD1 system will also significantly reduce our total cost of operation due to its higher density, superior heat dissipation and lower power consumption", Jan Odegard stated. "The standards-based Linux operating system allows us to provide a common hardware and software platform that enables collaboration and resource sharing among our various research groups, resulting in greater efficiency and economy.”

The acquisition of the supercomputer was funded by a $2 million federal grant, one of the largest awarded under the National Science Foundation’s Major Research Infrastructure programme. The system will initially support 32 faculty investigators and 150 graduate and undergraduate students. CITI anticipates the supercomputer will eventually benefit any Rice faculty member whose research depends on large-scale computing. Projects currently scheduled to run on the system include models of flows in implanted blood pumps, studies of thermal convection in the Earth's mantle and computer-aided drug discovery.

The Cray XD1 supercomputer leverages Direct Connect Architecture, the AMD Opteron HyperTransport technology and innovative memory controller features to provide high-bandwidth, low-latency links between processors and memory. This high-speed interconnect significantly accelerates applications, allowing users to tackle larger problems and solve them more quickly.

"We're excited that Rice University has selected the Cray XD1 supercomputer as the centerpiece for its research computing programme”, stated Peter Ungaro, president and chief executive officer of Cray. "Rice has an excellent reputation around the world, and we are energized by the opportunity to build our largest Cray XD1 supercomputer for their very demanding scientific applications. We believe it provides further proof of the value of Cray's innovative research and development in providing some of the world’s most advanced computing systems.”

"We designed Dual-Core AMD Opteron processors to help our customers realize more computing power and efficiency”, stated Ben Williams, vice president of commercial business at AMD. "Cray's XD1 supercomputer takes full advantage of AMD's true dual-core technology and Direct Connect Architecture, providing customers with systems that excel at some of the most demanding applications in use today."

The Cray XD1 supercomputer combines direct-connect system architecture, high performance computing (HPC)-optimized Linux, management and reconfigurable computing technologies to deliver exceptional performance on real-world applications. Purpose-built for demanding HPC applications such as computational chemistry, environmental forecasting and computer-aided engineering, the Cray XD1 system lets users simulate, analyse and solve complex problems more quickly and accurately. The AMD64-based Cray XD1 system supports a broad range of 32- and 64-bit HPC applications on AMD Opteron single- or dual-core processors. The Cray XD1 system also provides application acceleration capabilities using FPGA technology tightly connected to the direct-connect structure.

The Rice research system will support an expanding community of researchers in fields as diverse as biotechnology, nanotechnology, psychology, earth sciences, fluid dynamics and computer science. Designed to support hundreds of users in the future, the system will initially be used for a series of memory-intensive applications sponsored by the National Science Foundation.

For example, bioinformatics researchers will use the system to leverage robotic motion-planning algorithms for computer-aided drug design. Earth scientists will model and simulate, with great detail, the deformation of sediments, soils and other materials near the earth's surface. Psychologists will use the system to better understand the structure and functions of the human brain and how they impact the development of speech, language, memory, perception and motor skills. Computer scientists will test new programming tools, compilers and system software for high-performance computers. And fluid dynamics researchers will use the supercomputer to model how heart pumps and other devices might aid blood flow.