Accelerating Temporal Triangle Counting and Betweenness Centrality on GPUs

Temporal graphs have become very relevant and important in diverse and rapidly changing environments, such as in social network analysis, supply-chain management, urban planning, etc. In this work, we consider temporal graphs and propose GPU-based implementations of two graph algorithms that have widespread use, namely, Triangle Counting and Betweenness Centrality. We show that the GPU parallel implementations are very effective, providing significant speed-up when compared to their CPU counterparts.