What to Do with All That Bandwidth? GPUs for Graph and Predictive Analytics

Originally published at: https://developer.nvidia.com/blog/gpus-graph-predictive-analytics/

Figure 1: Graph algorithms exhibit non-locality and data-dependent parallelism. Large graphs, such as this map of the internet, represent billion-edge challenges to existing hardware architectures. Did you see the White House’s recent initiative on Precision Medicine and how it is transforming the ways we can treat cancer? Have you avoided clicking on a malicious website based…

Increasing memory bandwidth with GPU technology -- just the ticket for all the intensive analytics coming down the pike.

Thanks for a great overview of the graph analytics using GPUs. The article is quite informative and summarizes quite in detail the present state of affairs in GPU-based graph analytics.
I personally am trying to explore a research topic in the domain of graph-based predictive algorithms especially in the domain of predicting the spread of epidemic in a region. I intend to use CUDA to simulate the results. Can anyone provide an insight into the suitable areas in which one can research in this regard.