Brian Stinar

Senior majoring in Computer Science and minoring in History

What's YOUR college job like?

“My research involved creating energy-aware scheduling algorithms that performed task duplication on clusters of computers.

“A translation of that for non-computer scientists is as follows: if a computer has a problem to work on, that work can be broken up into tasks. Certain tasks need to finish before other tasks can begin. For scientific problems, many computers are able to work together; we call this system of computers a cluster.

“One way to make a cluster work quicker is to have a task solved twice on different computers within the cluster. This may seem counterintuitive, because we are wasting time recalculating the results of the task, but actually we are able to eliminate the communication cost between the different computers in the cluster. This technique works well in situations where communication is expensive.

“Our group further expanded this concept to take into account energy consumed by the cluster of computers. Using algorithms we developed, we are able to trade processing time for energy savings. This trade can be extremely useful. Large computing facilities, especially during rolling blackouts, would be a perfect place to apply these algorithms.

“The above idea was developed by my research group leader, Dr. Xiao Qin [New Mexico Tech Department of Computer Science]. Ziliang Zong, a Tech Ph.D. student created our algorithms. A fellow undergraduate student (now graduated) Adam Manzanares wrote the simulator we used to test our ideas. I modeled power consumption and helped with getting results from the simulator.”

Brian, a 2002 graduate of Cliff High School, created courses in “Introduction to Paintball” and “Competition Paintball”, both of which are very popular. He also represents New Mexico Tech paintball throughout the state on a traveling team.