Presenter: Ghadeer Alabandi
Advisor: Dr. Martin Burtscher
Date/Time: Wednesday, May 19 @10:00 AM (CDT)
Zoom Link: https://txstate.zoom.us/my/martin.burtscher
The exponential increase in the volume of data humans produce has prompted the data analytics field to grow rapidly over the last decade. In particular, the interest in analyzing and extracting information from large graphs has risen sharply. However, the extensive size and high complexity of some graphs incur a heavy computational cost, making it hard to study them using traditional analytical frameworks. Moreover, many important graph algorithms are difficult to parallelize, particularly for GPUs, because of their irregular memory accesses, data-dependent behavior, load imbalance, and memory consumption. The goal of my research is to tackle these challenges by developing new and better ways to accelerate important graph algorithms. To date, my contributions include 1) increasing the parallelism of graph coloring, 2) improving the quality of graph coloring, and 3) speeding up the detection of inequity in social networks.
Deadline: June 20, 2021, midnight