Last month, IBM announced the latest model in its supercomputer product line: the Blue Gene/P. The previous version, the Blue Gene/L, held the distinction of being the “fastest computer on earth,” clocking in at approximately 300 teraFLOPS (trillion floating point operations per second). The new generation system, Blue Gene/P, is designed to achieve speeds in excess of one petaFlop (1,000 teraFLOPS). To put that into perspective, Blue Gene/P is 100,000 times more powerful than a home PC and can process more operations in one second than the combined power of a stack of laptop computers nearly 1.5 miles high.
The one-petaFLOP Blue Gene/P configuration is a 294,912-processor, 72-rack system harnessed to a high-speed optical network. The Blue Gene/P system can be scaled to an 884,736-processor, 216-rack cluster to achieve a three-petaFLOP performance rating. A standard Blue Gene/P supercomputer configuration will house 4,096 processors per rack, using four IBM PowerPC 450 processors integrated on a single Blue Gene/P chip.
Energy and Space Efficiency
The Blue Gene/P was specifically designed to fit in small spaces and use less electricity than other commercially-available designs. As a result, IBM states that the system is at least seven times more energy efficient than any other supercomputer.
Some of the world’s leading research laboratories and universities have already placed orders for Blue Gene/P supercomputers. The U.S. Department of Energy’s Argonne National Laboratory, Argonne, Ill., will deploy the first Blue Gene/P supercomputer in the U.S. later this year. In Germany, the Max Planck Society and the Forschungszentrum Julich also plan to install Blue Gene/P systems in late 2007. Additional Blue Gene/P system rollouts are being planned by Stony Brook University and Brookhaven National Laboratory in Upton, New York and the Science and Technology Facilities Council, Daresbury Laboratory in Cheshire, England.
At FZ Julich, where researchers have been using a Blue Gene/L machine for two years, the Blue Gene/P system is expected to help advance more breakthroughs in fields such as particle physics and nanotech while keeping the research facility within acceptable power budgets. “The big computing power at low electricity rates allows us to boost the performance of very complex and computationally intensive algorithms,” said Thomas Lippert, director of the supercomputing center at FZ Julich.
A Friendlier Application Development Environment
There are some key differences between the new Blue Gene/P and its earlier counterpart. In terms of hardware, the Blue Gene/P supercomputer has more (four vs. two) and speedier (850 MHz vs. 700 MHz) processors per chip, more memory, and an SMP mode to support multi-threaded applications. This new SMP mode moves the system to a programming environment similar to that found in commercial clusters. Additionally, system management has been significantly enhanced and simplified in the new system.
The Blue Gene/P operating system is based on the open-source Linux operating system. Applications are written in common languages such as Fortran, C, and C++ using standards-based message-passing interface (MPI) communications protocols. The supercomputer is compatible with the diverse applications currently running on the Blue Gene/L system, including applications in physics, chemistry, biology, aerospace, astrophysics, genetics, materials science, cosmology and seismology.
In addition to its traditional uses, IBM expects that new commercial applications will take advantage of the power and more user-friendly environment offered on the system. Dave Turek, Vice President of Deep Computing at IBM said, “A new group of commercial users will be able to take advantage of its new, simplified programming environment and unrivaled energy efficiency. We see commercial interest in the Blue Gene supercomputer developing now in energy and finance, for example. This is on course with an adoption cycle–from government labs to leading enterprises–that we’ve seen before in the high-performance computing market.”
While there may be interest from some commercial users in such industries as energy and finance, it will be limited to the largest of companies with massive processing requirements. However, if IBM does manage to break into the commercial market with the Blue Gene/P, it will likely mean that its successor will have even more commercially-appealing design features.