Professor's Research Team Wins International Supercomputing Scaling Contest
- Friday, Jul 01, 2011
An innovative application of cloud computing to enhance oil recovery won first place for researchers at The Institute for Computational Engineering and Science's (ICES) Center for Subsurface Modeling who competed in the international contest highlighting best methods in supercomputing scaling.
The IEEE SCALE 2011 Challenge selected the multi-institutional team from The University of Texas at Austin, Rutgers University and IBM for the honor.
The winning team started with UT's successful enhanced oil recovery supercomputer software that models underground aquifers developed by researchers under Mary F. Wheeler, director of the ICES Center for Subsurface Modeling and professor of aerospace engineering, engineering mechanics, petroleum and geosystems engineering, and mathematics. Wheeler also holds the Ernest and Virginia Cockrell Chair in Engineering.
With a goal to make the effective oil recovery software easier to use and available on demand, researchers from Rutgers University and IBM created software interfaces and filters to make it accessible through cloud computing.
Using an iPad as the interface, their prize-winning project demonstrated how cloud computing, with its easy-to-use features, effectively supported extremely powerful, widely geographically separated supercomputers. The multi-institutional, multi-national team accessed two high-end supercomputers — one located at KAUST in Jeddah, Saudi Arabia, and the other at IBM's research center in New York.
Gergina Pencheva, research associate, and Reza Tavakoli, postdoctoral fellow, were Wheeler's other team members who worked with groups from The Center for Autonomic Computing at Rutgers University and IBM's T.J. Watson Research Center.
SCALE 2011, the Fourth IEEE International Scalable Computing Challenge, highlights and showcases real world problem solving using computing that scales. The contest focused on end-to-end problem solving using concepts, technologies and architectures that facilitate scaling.