A team of UCI and Dartmouth students study the mind's problem-solving skills with a cluster of Playstation 3 systems.
University of California Irvine students Jayram Moorkanikara, Jeff Furlong and Matt Johnson, along with two Dartmouth College partners, ranked among the first-place winners of IBM's Cell Broadband Engine Professor University Challenge on Sept. 24 by using three Sony Playstation 3 systems to recreate how the human brain functions. The team's project, which was showcased at the 2007 Power Architecture Developer Conference in Austin, Texas, centered on reproducing brain algorithms, procedures performed by the brain to complete everyday functions, and visual processing, which has traditionally required complex computers to perform efficiently.
Moorkanikara disucssed how he was inspired for this project by the intricacies of the mind. "I was interested in understanding and building hardware that mimics the behavior of the brain and that can be useful for many diverse applications. Understanding the features and ability of the brain have inspired me to look at these algorithms and develop suitable hardware to implement them.
"Most of the gaming systems ... are totally designed to be best in games only. I felt that they are like a race car designed for driving on race tracks only [and are] not effective on normal roads. Other gaming systems' hardware are customized for graphics used in gaming applications.
"PS3 contains the IBM CELL processor, whose architecture is suitable for parallel programming and can be used not only for games but for many scientific and multimedia applications. We have also demonstrated that a cluster of these PS3 can behave like a mini supercomputer."
The UCI/Dartmouth team received a $10,000 award after competing against 80,000 other college students. Despite their success so far, Moorkanikara doesn't see them stopping. "We have just started looking into the exciting area of hardware realization of brain algorithms. More such algorithms and applications will be developed and used in smart, self-guided cars, computers and robots."
Source: New University