SGI used to predict Ivan

National Oceanic & Atmospheric Administration

When NASA weather forecasters tried to plot the path of Hurricane Ivan this week, they turned to a Silicon Graphics Inc. supercomputer.

"The actual prediction of events like hurricanes is one of the hardest problems in supercomputing," said Jill Matzke, SGI high performance computing marketing manager. "It's especially difficult due to the time constraints."

Supercomputers are coveted for their vast simulation capabilities, from predicting weather to simulating space missions. SGI's supercomputers helped weather forecasters analyze the path, direction, intensity and rainfall amounts of Hurricane Ivan.

NASA's SGI machine uses 5,000 processors. SGI officials also provided supercomputers to the National Oceanic and Atmospheric Administration's Geophysical Fluid Dynamics Laboratory (GFDL) in Princeton, N.J., which run the weather models that are helping the nation to prepare for hurricanes.

NASA officials previously announced a partnership with Silicon Graphics and Intel Corp. at Ames Research Center to build a super fast machine as part of Project Columbia. This project, Altix, is Linux-based with more than 10,000 processors.

This summer, the House passed the Energy Department High-End Computing Revitalization Act of 2004, which established a research and development program within Energy to develop more advanced computers. House officials also approved the High Performance Computing Revitalization Act of 2004, which seeks to recapture U.S. dominance in supercomputing by requiring the Energy secretary to carry out an R&D program

In late September 2000, NOAA officials awarded Raytheon Co. a three-year, $34 million base contract to build the high-performance computing system. The total value of the contract, inclusive of all options, is approximately $67 million. Raytheon officials selected the SGI Origin 3000 series to upgrade GFDL's supercomputing capabilities.