Louisiana State University Selects SGI Technology for Storm Modeling and Scientific Visualization

Silicon Graphics Prism Visualization System Aids Hurricane Katrina Recovery

Efforts; SGI Technology to Be Available to Eight Colleges and Universities

via 40 Gb Louisiana Optical Network Initiative

Mar 08, 2006, 00:00 ET from SGI

    MOUNTAIN VIEW, Calif., March 8 /PRNewswire-FirstCall/ -- To enable a wide
 range of scientific discovery efforts, including storm surge mapping of
 Hurricane Katrina, Louisiana State University (LSU), in August 2005, installed
 a powerful combination of technology from Silicon Graphics
 (OTC Bulletin Board:   SGID) for its Center for Computation & Technology.
 Purchased through James River Technical, Inc., SGI exclusive higher education
 reseller, the Silicon Graphics Prism(TM) Extreme visualization system and
 SGI(R) Visual Area Networking (VAN) technology with SGI(R) OpenGL
 Vizserver(TM) software will also add real-time 3D collaborative visualization
 to the emerging statewide 40Gb optical network called LONI (Louisiana Optical
 Network Initiative). LONI, which is connected to the National LambdaRail
 (NLR), a nationwide optical network infrastructure, will directly link the
 Silicon Graphics Prism(TM) visualization system at LSU with eight Louisiana
 schools including Tulane and Xavier Universities, and other smaller
 universities. The SGI technology purchase was made possible by grants from the
 National Institutes of Health (NIH) and the National Science Foundation (NSF).
     The Silicon Graphics Prism visualization system, called "Santaka," after a
 spicy Louisiana pepper, and SGI Vizserver technology join a 1,024-processor
 Linux(R) cluster and other similar clusters in LSU's High Performance
 Computing group, a joint partnership between the Center for Computation &
 Technology (CCT) and Information Technology Services (ITS), on the main campus
 in Baton Rogue. When considering the advent of LONI, and the growing
 computational demands in many fields, especially biology, biochemistry,
 computational chemistry, physics, and computational fluid dynamics, LSU's
 Director of High Performance Computing (HPC) decided the time was right to
 bring in another system and architecture.
     "After initially considering several other systems and the SGI Altix
 servers, I met with others in the research community here on campus to discuss
 their requirements," said Brian Ropers-Huilman, Director of HPC, LSU. "I had
 quotes from several vendors, we discussed it openly, and in the end there was
 much more interest in the Altix, in terms of it having a shared-memory system
 rather than distributed memory like our clusters. With the size of our
 datasets rapidly growing, everyone believed that having substantial shared
 memory would enable us to conduct more and better research in less time.
 People were also quite interested in the visualization and graphics capability
 that came with the name SGI. That took us down the Silicon Graphics Prism
 visualization path, which was extremely appealing."
     Simulating Katrina's Storm Surge
     There are a number of storm modeling groups on the LSU campus, loosely
 tied together under one umbrella. One of the major groups is the Center for
 the Study of Public Health Impacts of Hurricanes (CSPHIH), led by Dr. Ivor van
 Heerden, a hurricane expert frequently interviewed on CNN and other news
 broadcasts. Before, during, and in the aftermath of Hurricane Katrina, CSPHIH
 computed the ADCIRC tidal surge model on LSU's 1024-processor SuperMike Linux
 cluster and then viewed the data in 2D on small workstations. The group
 regularly provides 2D data to Emergency Operations Centers (EOCs) in Baton
 Rouge as well as other states, when hurricanes are bearing down on the coast.
 With the 2006 Hurricane Season less than three months away and already
 expected to be about as active as 2005, the Silicon Graphics Prism
 visualization system is expected to improve the way data is provided to EOCs.
     The Visualization, Interaction, and Digital Arts (VIDA) division of CCT
 took this and other simulations and observational data, and viewed them in 3D
 on the Silicon Graphics Prism visualization system. Underlying datasets
 include the MM5 atmospheric model (including temperature gradients and wind
 field vector data); satellite infrared time-series images; and tidal surge
 data. Using multi-resolution aerial photography of New Orleans, their
 visualization zooms into the Louisiana coast in a spectacular fly-in opening,
 right through the eye of the hurricane, soon followed by the tidal surge and
 its flooding aftermath.
     "Our Katrina efforts provide both a compelling stereo fly-through, as well
 as a collaboratively-manipulable, fully interactive visualization of Katrina,"
 said Dr. Brygg Ullmer, assistant professor, LSU CS + CCT, and leader of VIDA's
 visualization group. "Several of these datasets are large-scale simulations,
 seeded by actual measurements of the storm; others include data from sensors
 and satellites. We've transformed these into two compelling kinds of visuals.
 One of these is a visual flythrough, viewable in stereo with inexpensive
 glasses. The second is a fully interactive visualization, which allows
 multiple participants to collaboratively steer time and other parameters
 through new 'tangible' interaction devices. We're now trying to leverage that
 interactivity, hosting it on the Prism system so that people all over
 campus -- or throughout the state -- can inspect, manipulate, and dissect the
 anatomy and physiology of Katrina. The Silicon Graphics Prism's shared memory
 and VizServer software are key enabling tools toward this. We hope this can
 become a compelling research tool and outreach vehicle for many people across
 the state -- whether they're studying the weather or ecological issues,
 planning for the next hurricane, or just trying to better appreciate the
 enormity of the event."
     "Hurricane Katrina's devastation has provided a significant opportunity
 for scientific research and simulation -- all focused on avoiding similar
 disasters in the future. The models needed to simulate its effects under
 differing conditions in real time require the processing of enormous data sets
 and the collaboration of researchers throughout Louisiana and the country.
 These collaborations, with the aid of the Louisiana Optical Network Initiative
 and the National Lambda Rail, are made possible with the industry-leading
 shared memory capabilities of the SGI Extreme Prism, and the ease of its
 integration into a heterogeneous distributed cluster environment, like that at
 LSU," said Tom Mountcastle, President of James River Technical.
     LSU purchased the Silicon Graphics Prism visualization system with six
 ATI(R) FireGL(TM) graphics processors, 32 Intel(R) Itanium(R) 2 processors
 running Linux(R) OS with SGI ProPack 4 for Linux, and a 10gigE interface as
 well as SGI OpenGL Vizserver software.
     SILICON GRAPHICS | The Source of Innovation and Discovery(TM)
     SGI, also known as Silicon Graphics, Inc., is a leader in high-performance
 computing. SGI helps customers solve their computing challenges, whether it's
 sharing images to aid in brain surgery, finding oil more efficiently, studying
 global climate, providing technologies for homeland security and defense,
 enabling the transition from analog to digital broadcasting, or helping
 enterprises manage large data. With offices worldwide, the company is
 headquartered in Mountain View, Calif., and can be found on the Web at
     NOTE:  Silicon Graphics, SGI, Altix, the SGI cube and the SGI logo are
 registered trademarks and Silicon Graphics Prism and The Source of Innovation
 and Discovery are trademarks of Silicon Graphics, Inc., in the United States
 and/or other countries worldwide. Intel and Itanium are trademarks or
 registered trademarks of Intel Corporation or its subsidiaries in the United
 States and other countries. Linux is a registered trademark of Linus Torvalds
 in the U.S. and other countries. All other trademarks mentioned herein are the
 property of their respective owners.
      Marla Robinson