Two Cray Supercomputers to Surpass the 100 Teraflops Mark

Cray's XT-Based Systems Feature a Scalable, Upgradable Architecture to Meet Massive Computational Demands While Leveraging Customer Investments: Cray today announced that two Cray supercomputers are now in an elite group of systems that can perform computations at more than 100 teraflops (100 trillion floating point operations per second), as measured by the industry-standard TOP500 benchmark. The Cray systems, one installed at Oak Ridge National Laboratory (ORNL) and the other at Sandia National Laboratories, are now two of the three fastest computers in the world. They are built on a Cray XT infrastructure that enables Cray customers to upgrade to increasingly higher performance levels, instead of forcing them to invest in an entirely new system. The Cray XT4 supercomputer at the Department of Energy's ORNL, nicknamed Jaguar, is the world's most powerful system dedicated to open science. ORNL completed an upgrade to its existing Cray XT system earlier this year that more than doubled its performance to over 100 teraflops. When running the nuclear fusion application AORSA, the ORNL system achieved an impressive 74 percent of the system's theoretical peak, an efficiency level that usually can be attained only when performing "artificial" benchmark tests. The system has at least one more major upgrade in its future that will double its performance once again. The Cray XT3 supercomputer at Sandia, known as Red Storm, was developed jointly by Cray and Sandia, a part of the Department of Energy's National Nuclear Security Administration. The system operates in a dual network environment that allows for both classified and unclassified work. Sandia upgraded Red Storm late last year to three times its original performance level, boosting its performance to more than 100 teraflops. "We want to thank and congratulate both Oak Ridge and Sandia for their achievement in reaching this impressive performance milestone," said Cray President and CEO Peter Ungaro. "Typical Linux clusters can't scale to these performance levels, especially on real-world applications," continued Ungaro. "And other custom systems are based on special microprocessors and packaging technologies that cannot cost-effectively be upgraded to take advantage of new technologies. The Cray XT design combines AMD Opteron processors and a custom Cray SeaStar network to provide unparalleled scalability and upgradeability. This allows our customers to leverage their investment in both hardware and software by increasing the size and capabilities of their existing supercomputers as the technology advances, rather than having to purchase a new system. It's a big total cost of ownership (TCO) win for them. In fact, perhaps the best proof point of this TCO advantage is that more than 75% of our XT customers have upgraded their initial Cray system."
Like
Like
Happy
Love
Angry
Wow
Sad
0
0
0
0
0
0
Comments (0)