Shattering expectations for real-time GPU and GPUDirect visualization - On-Demand GTC Session # SS33029
Only a few visualization tools take full advantage of GPUs and GPUDirect to deliver real-time graph analytics across petabytes of incoming and stored data. Graphistry’s breakthrough software and pipeline technology, combined with NVIDIA GPUs and GPUDirect along with Pavilion’s universally unmatched storage enables the next generation of interactive GPU visualization for incident response, SIEM optimization and threat hunting across market segments including financial services, federal agencies and data science.
Please check this on-demand session with Pavilion and Graphistry at GTC21, to learn about Visually Investigating Patterns in Logs at Scale with Graphistry, RAPIDS, and Pavilion using the most performant, dense scalable, and flexible storage platform in the universe.
We share the applications and architectures that deliver the best results. We also describe combining components: UMAP/cuGraph pattern mining, cuDF data wrangling, Graphistry/Jupyter/StreamLit point-and-click UI, then add in dask-cudf, GPUDirect, and Pavilion storage.
Finally, we discuss what is easy/hard to do and provide recommendations. Viewers can run and fork the open-source reference architecture and chat about it with the RAPIDS community.If you want to learn more about how Pavilion is shattering expectations for GPU-based workloads, check out these on-demand sessions and blogs:
How to respond to rapidly scaling Geospatial Intelligence with OmniSci and Pavilion
How a multi-controller storage architecture shatters expectations for modern applications
