Out of 17,000 applicants from across the country, Computer Science senior Ben Brock won the prestigious National Science Foundation (NSF) Graduate Research fellowship for his idea to research parallel computing that can run on graphics processors (GPUs).
This five-year fellowship allows Ben the academic freedom to pursue his research interest in the Ph.D. program of UC Berkeley, where he expects to continue working on new programming models for using new computer architectures.
Born in Hong Kong, Brock moved with his family to Talbott, Tennessee when he was three years old. When he was 8, they relocated to Macau for three years before coming back to Tennessee and settling in Jefferson City.
Ben chose Computer Science because he has always loved computers, and because he had enjoyed programming as a self-taught hobby since middle school. “I really liked the idea of hooking up computers together to make supercomputers, so I dabbled a little bit with hooking up my computers to make a computer cluster in my bedroom at home.”
Upon graduating from high school, Brock decided to stay in Tennessee for college. “I chose UT mostly because it offers the resources of a big university, but also some programs, like the Haslam Scholars Program, that offer the fellowship and community of a liberal arts college.”
In addition to his CS studies, Ben serves as chairman of the Honors Community Board, freshman representative to the Honors Council, and a staff writer for the satirical newspaper The Tangerine. An avid alto saxophonist, Ben was first chair in his high school concert band and was selected to join the Tennessee All-State East Red Band as a sophomore. His hobbies include computer programming, jogging, music composition, and reading.
About winning this fellowship, Ben said, “It’s really nice, in terms of funding, and in being able to do whatever kind of research you want to do, because you don’t have to have a particular professor who’s going to fund you. You can just work on whatever research you like.”
Brock wants to research methods for parallel computing. “I am currently working with a professor from the physics department on a computational physics problem, where we look at burning. We put a bunch of chemicals in a box and burn them, and see what happens as they react to each other, after a certain amount of time. That’s called chemical kinetics, chemicals interacting with each other, and it’s applicable to a lot of fields, but we’re specifically looking at explosions in supernovas. I am working on the code for evaluating those chemical kinetics, and figuring out what happens to those chemicals during supernova explosions,” he said. This code, called Fast Explicit Reaction Networks, or FERN, is very fast, runs in parallel and can run on graphics processors of the type that run a computer’s display. Over the past five years or so, researchers have developed programming models to use GPUs for general purpose computing.
What Brock is interested in, having done “a lot of science hand-tuning to get things working,” is “coming up with new programming models for using new computer architecture without making the programming really annoying to have to deal with.” He wrote his proposal on the topic and submitted it as part of the fellowship application process.
“The Graduate Research Fellowship Program is a vital part of our efforts to foster and promote excellence in U.S. science, technology, engineering and mathematics by recognizing talent broadly from across the nation,” said Joan Ferrini-Mundy, NSF assistant director for Education and Human Resources. “These awards are provided to individuals who have demonstrated their potential for significant research achievements, and they are investments that will help propel this country’s future innovations and economic growth.” More information is available in the GRFP announcement by NSF.