Computers are fast. They benefit humanity because of their ability to process data much faster than a human can. But Gokarna Sharma, Ph.D., assistant professor in the Department of Computer Science in ɫҹ University’s College of Arts & Sciences, thinks they can be faster, and now he has additional funding to teach computers - and students - how to make that happen.
Sharma recently received a CAREER Award from the National Science Foundation (NSF) for research and technological advancement. This award is primarily given to early-career faculty within the first 6 years of their academic career who show outstanding potential to serve as academic role models in research and education and to lead advances in the mission of their department or organization.
Sharma received his funding to implement a system dedicated to increasing the rate that tasks can be completed within a computing system.
This grant will also allow Sharma to mentor K-12, undergraduate and graduate students. This grant provides funding for two graduate students for five years so that they can assist Sharma on research. This grant will help him support undergraduate research activities through the Division of Research & Sponsored Programs’ Summer Undergraduate Research Experience (SURE) program at ɫҹ and K-12 research activities through the Science Experience program at ɫҹ.
Sharma’s goal is to improve the mechanisms responsible for minimizing the time it takes for computers to complete tasks. Sharma is working to develop a system skilled enough to complete many orders simultaneously by creating algorithms advanced enough to speed up the process. Sharma will exploit the novel concept of “transactional memory” in creating such algorithms. This novel concept allows non-conflicting tasks to complete in parallel and conflicting tasks will be completed in time no more than that is needed while not using the concept.
“Lots of computer tasks are stacked up one after the other,” said Sharma. “If we can create algorithms that complete multiple tasks concurrently even in the situations that the tasks may share data, it will increase the response rate for a vast number of systems. Multiprocessor systems are being used everywhere, and any technique that simplifies programming those systems will eventually help the general public.”
Faculty only get three chances to win an NSF CAREER award which is one of the reasons that these are so competitive. Sharma’s first submission helped him understand what the review panel was looking for within the proposal. His second draft was in much better condition, but still had some loose ends to be tied up. His third submission was approved by the panel.
“While a tedious process, it was such a great learning experience,” Sharma said. “It doesn’t matter how much you know about your field of study.
“The important thing is getting the review panel excited about your research in a comprehensible way. Writing it from the perspective of a reviewer helped me receive this grant.”