Abstract
We show that dynamic graph algorithms are amenable to parallelism on graphics processing units (GPUs). Evolving graphs such as social networks undergo structural updates, and analyzing such graphs with the existing static graph algorithms is inefficient. To deal with such dynamic graphs, we present techniques to (i) represent evolving graphs, (ii) amortize the processing cost over multiple updates, and (iii) optimize graph analytic algorithms for GPUs. We illustrate the effectiveness of our proposed mechanisms with three dynamic graph algorithms: dynamic breadth-first search, dynamic shortest paths computation and dynamic minimum spanning tree maintenance. In particular, we show that the dynamic processing is beneficial up to a certain percentage of updates beyond which a static algorithm is more efficient.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
