GTC 2020: High-Performance, Remote Scientific Visualization in Jupyter Notebooks
After clicking “Watch Now” you will be prompted to login or join.
Click “Watch Now” to login or join the NVIDIA Developer Program.
High-Performance, Remote Scientific Visualization in Jupyter Notebooks
Nick Leaf, NVIDIA | Maximilian Rietmann, NVIDIA
We'll introduce iPyParaView, a Jupyter widget for interactive rendering in the notebook with ParaView. iPyParaView leverages the GPU in your local or remote Jupyter instance to render data without copying it back to the client. It uses ParaView's Python bindings to expose its full capabilities — including the RTX path-tracing back end and IndeX volume rendering plugin. We'll use live demos to show how to pre-process data with RAPIDS and render with ray-tracing, as well as how to use a Dask-MPI cluster for distributed-memory processing and visualization.