Andrew Belliner

Graphics Processor-Based Implementation of Bioinformatics Codes

We created a powerful computing platform based on graphics cards to improve the performance of bioinfo- matics codes. A graphics processing unit (GPU) is a component of a computer responsible for showing a display on the screen. To keep up with the video gaming industry, modern GPUs have become very advanced computational devices. Recently computer scientists have taken interest in a GPU's ability to perform computational commands on multiple sets of data simultaneously, leading to improved performance.

Our system contains two GPUs, capable of working in parallel, for a total of 256 processors (for comparison, a modern desktop will have two or four processors). Our plan includes using this computational power to improve the performance of existing bioinformatics codes, specifically, the Smith-Waterman algorithm for finding DNA or protein sequence alignment.

Faculty Mentors: Greg Wolfe and Christian Trefftz, Computer Science

UPDATE:

Andrew took First Place at the Student Research Competition of the Consortium for Computer Science in Colleges Midwest Conference Sept. 27-18, 2008.

Andrew presented at the ACM/IEEE Supercomputing Conference November 15-21, 2008 in Austin, TX.

Andrew can be seen on YouTube talking about his project.

Page last modified July 14, 2009