Publication Type

Conference Proceeding Article

Publication Date

10-2017

Abstract

As graph analytics often involves compute-intensive operations,GPUs have been extensively used to accelerate theprocessing. However, in many applications such as socialnetworks, cyber security, and fraud detection, their representativegraphs evolve frequently and one has to perform arebuild of the graph structure on GPUs to incorporate theupdates. Hence, rebuilding the graphs becomes the bottleneckof processing high-speed graph streams. In this paper,we propose a GPU-based dynamic graph storage schemeto support existing graph algorithms easily. Furthermore,we propose parallel update algorithms to support ecientstream updates so that the maintained graph is immediatelyavailable for high-speed analytic processing on GPUs. Ourextensive experiments with three streaming applications onlarge-scale real and synthetic datasets demonstrate the superiorperformance of our proposed approach.

Discipline

Computer Engineering | Data Storage Systems

Publication

Proceedings of the 44th International Conference on Very Large Data Bases, Rio de Janeiro, Brazil, 2017 October

Volume

11

ISBN

2150-8097

City or Country

Rio

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS