Dr. Bruce Lester, Maharishi University of Management Professor of Computer Science, received the Best Paper Award of The 2011 International Conference of Parallel and Distributed Computing held in London, U.K., July 2011. This conference was part of the annual World Congress on Engineering sponsored by the International Association of Engineers.
Dr. Lester’s paper “Improving the Performance of Collection-Oriented Operations through Parallel Fusion” presents a technique for utilizing multiple computer processors to make programs run faster.
Dr. Lester founded the Computer Science Department at Maharishi University of Management and served as Chair for eight years. He has six years of experience as a Senior Software Engineer developing client/server database systems and internet websites for Primus Telecommunications, an international long-distance telephone company. His teaching experience includes two years at Princeton University and eleven years at Maharishi University of Management.
Dr. Bruce Lester
A brief explanation of his paper
All new computers today have processors with multiple “cores.” Each core is essentially a separate processor capable of executing its own stream of instructions in parallel with other cores. Increasing the number of cores per processor improves the performance of the computer. The number of cores per processor is expected to double every three years. This offers the potential for greatly improved computer performance.
International Association of Engineers
However, this potential of the computer hardware is not yet fully realized by the software. One promising technique for solving this problem is called data parallel programming. In data parallel programming, the same task is applied to different aspects of the data collection process simultaneously. In the ordinary implementation of data parallel operations, all the cores must synchronize with each other after each task. In his paper, Dr. Lester presents a technique for removing this time-consuming synchronization, thus speeding up the execution by as much as 80%.
Want more details? Check out the complete paper.
 |
Anna Bruen is a MUM alumnus who graduated with a degree in Sustainable Living. Anna is the resident blogger here at MUM. To learn more about Anna take a look at her introduction post. |
{ 2 comments… read them below or add one }
is their a limit to the number of core that can be available to commercially sound computers?
Anna, thanks for posting a link to Dr. Lester’s paper.
Alan