Cray Logo
spacer

blog  facebook  twitter  linkedin  google plus  youtube
graphicHomeSupportCustom EngineeringIndustry SolutionsProgramsAbout Cray
graphicHome > Investors > News Release
graphic
spacer
graphic
 
graphicInvestors
graphic
Overview
graphic
News Releases
graphic
Proxy Materials
graphic
Presentations and Webcasts
graphic
Investor Events
graphic
graphic
SEC and Section 16 Filings
graphic
Corporate Governance
graphic
Request Information
graphic
 

News Release

Printer Friendly Version View printer-friendly version
<< Back
Scientists Use Powerful Cray Supercomputer to Develop Groundbreaking Strategies in Weather Prediction
Latest Computer Models Zoom Down to Level of Individual Storm Cells

SEATTLE, WA, Aug 29, 2007 (MARKET WIRE via COMTEX News Network) -- Global supercomputer leader Cray Inc. (NASDAQ: CRAY) today announced that scientists have leveraged the power of a Cray supercomputer at the Pittsburgh Supercomputing Center (PSC) to break new ground in weather prediction. Researchers from the University of Oklahoma's Center for Analysis and Prediction of Storms (CAPS) and the National Oceanic & Atmospheric Administration (NOAA) employed an innovative combination of high resolution and controlled manipulation of numerical model parameters and starting conditions to develop strategies that will allow forecasters to better anticipate the formation of severe storms and the supercells that give birth to destructive tornados. The research was part of the NOAA Hazardous Weather Testbed (HWT) Spring Experiment.

"Each day during the Spring Experiment that was conducted from mid-April through early June, CAPS scientists applied emerging scientific methods to design a 10-member 'ensemble' of forecasts from the Weather Research and Forecasting, or WRF, software model," explained Dr. Ming Xue, director of CAPS. "Each member had a 4-kilometer horizontal resolution and covered almost the entire continental U.S. Unlike a single-model forecast, this ensemble not only predicted when and where particular weather might occur, but also the likelihood of its occurrence."

"Ensembles have been employed by larger-scale weather models before, but they've never been focused on the few-kilometer scales where individual storms actually occur," Dr. Xue continued. "The ensemble approach is exceptionally demanding when it comes to computational power and can only be accomplished on a high-performance, scalable system such as the Cray XT(TM)-based system at PSC."

Every day during the course of the experiment, terabytes (trillions of bytes) of data were generated, archived and transferred from PSC to Norman, Oklahoma, for use in forecasts, evaluations and future analysis and research. The HWT facility in Norman is strategically located in the recently built National Weather Center between the operational forecast areas of the NOAA Storm Prediction Center (SPC) and the NOAA National Weather Service Norman Forecast Office. These two offices, together with the NOAA National Severe Storms Laboratory (NSSL), led the experiment during the time of year when severe storm activity typically peaks in the region.

"The researchers had previously struggled to complete a single modeling run per day thus hampering a comprehensive understanding of how severe storms and tornadoes form," said Per Nyberg, Cray Marketing Director for Earth Sciences. "The scalability and sustained performance of the Cray XT system at PSC allowed them to complete 11 runs each day while using more sophisticated parameterizations. This is a key step in helping forecasters predict violent storms in time to prevent injury and loss of life."

About the NOAA Hazardous Weather Testbed Spring Experiment

The Spring Experiment conducted jointly by the SPC and NSSL would not have been possible without contributions from multiple partners. These two organizations, along with NOAA's Environmental Modeling Center (EMC) and the National Center for Atmospheric Research (NCAR), worked with CAPS and PSC in the design and execution of the ensemble forecasts. In addition, EMC and NCAR provided separate high-resolution WRF model forecasts for a complementary portion of the experiment. The WRF model was developed primarily at NCAR and EMC. The CAPS forecasts were produced under the support of the NOAA CSTAR program and the National Science Foundation Linked Environments for Atmospheric Discovery Large ITR project. The NSF TeraGrid and National Lambda Rail networks connected the groups in Pittsburgh and Norman. Go to http://hwt.nssl.noaa.gov/Spring_2007/ for more information.

About Cray Inc.

As a global leader in supercomputing, Cray provides highly advanced supercomputers and world-class services and support to government, industry and academia. Cray technology enables scientists and engineers to achieve remarkable breakthroughs by accelerating performance, improving efficiency and extending the capabilities of their most demanding applications. Cray's Adaptive Supercomputing vision will result in innovative next-generation products that integrate diverse processing technologies into a unified architecture, allowing customers to surpass today's limitations and meeting the market's continued demand for realized performance. Go to www.cray.com for more information.

Cray is a registered trademark, and Cray XT is a trademark, of Cray Inc. All other trademarks are the property of their respective owners.

Cray Media:
Erin McGhee
206/701-2107
pr@cray.com

Cray Investors:
Vic Chynoweth
206/701-2094
ir@cray.com

SOURCE: Cray Inc.

 

   

Print Page Print Page | E-mail Page E-mail Page | RSS Feeds RSS Feeds | E-mail Alerts Investor Alerts | IR Contacts IR Contacts | Financial Tear Sheet Financial Tear Sheet

 

graphic