by CIO Staff

Stanford Taps Tech for Earth Sciences Research

News
Jun 20, 20062 mins
Data CenterEnterprise Applications

Stanford University Tuesday is opening the doors to a new research center that is tapping technology from Sun Microsystems to better understand Earth sciences.

“Our long-term goal is to build large-scale, big integrated models that do analysis, simulation, and to some degree predictions of what might happen in future,” said Jerry Harris, director of the new Stanford Center for Computational Earth and Environmental Science (CEES) and a professor of geophysics at Stanford.

The center is a joint collaboration among Stanford University, government agencies and private industry. Sun contributed hardware and software including its x64 Opteron-based servers, Sparc-based Sun Fire servers and Solaris operating system.

The systems are helping fuel research into such areas as climate change, earthquake science and oil exploration.

“Those models on desktop computers take days; running them on the new system takes a few hours. So we can ask more what-if questions and run more scenarios. We can start to develop some statistical understanding of how things might change in the future and how human interventions might change those things,” Harris said.

For example, the researchers are studying how the oceans take up carbon and how potential alterations to the oceans could affect the climate. They’re also using satellite-based radar to stream down petabytes’ worth of data to study images showing defamation to the Earth’s surface, like in the case of an earthquake or water withdrawal from the Earth’s subsurface in drought-ridden areas such as Las Vegas.

Right now the researchers match applications to computing resources based on memory needs, but they plan to work with Sun and other IT providers to further refine computer systems to suit scientific applications.

“Ultimately we want to integrate applications and develop bigger, multidiscipline applications, where some pieces of the application run on large memory systems and other pieces run on distributed systems and then bring together the results in a coupled way,” Harris said.

-Shelley Solheim, IDG News Service (New York Bureau)

Check out our CIO News Alerts and Tech Informer pages for more updated news coverage.