NVIDIA IndeX is a 3D volumetric visualization tool that allows scientists and researchers to visualize and interact with massive datasets, make modifications and navigate to the most pertinent parts of the data, all in real time, to gather better insights faster.
Starting today an IndeX container is available on NVIDIA GPU Cloud (NGC). Simply pulling the container, and you can run it on your NVIDIA Volta or Pascal-powered workstation, NVIDIA DGX systems or in the cloud.
NGC provides HPC containers that package the application with appropriate accelerated libraries, MPI and other dependencies. This lets researchers focus on tackling scientific challenges instead of waiting for a system admin to navigate through complex installation processes on a shared environment.
And NGC offers a slew of HPC visualization tools that can be easily deployed for volume rendering, ray tracing and analysis of scientific data.
We’ve added two new containers on NGC: VMD and NVIDIA IndeX. The newest version of IndeX has a host of new capabilities to that allow users to:
- Speed up the discovery process – CUDA programming interface allows rewriting the core data visualization routines in real time and gives instantaneous visual feedback, enabling users to experiment with different kinds of visualization interactively.
- Better justify observed phenomena – CUDA programmable data query operations offer the ability to probe the data, access and analyze the real physical data (such as entropy or velocity).
- Accelerate large-scale data visualization – using NVLink high-speed interconnect technology between GPUs to easily scale performance.
- Visualize the most complex data structures – using the NVIDIA OptiX ray-tracing API to build the acceleration structure for visualizing unstructured meshes on the GPU. The faster building process of the acceleration structures makes real-time in-situ visualization of complex data structures possible.
- Immediate access via HTML 5-based viewer – the viewer’s source code is part of the NVIDIA IndeX SDK.
NVIDIA IndeX works by leveraging GPU clusters for scalable, real-time visualization and computing of multi-valued volumetric data together with embedded geometry data.
Tackling Scientific Visualization with NVIDIA IndeX
At the GPU Technology Conference, taking place through March 29 in San Jose, we’re collaborating with researchers from UC Berkeley, the University of Illinois at Urbana and Caltech to demonstrate how IndeX can be used to interact with and analyze an HPC simulation of a supernova.
Scientists have studied supernovae for decades to learn about the properties of exploding stars and get better insights into the evolution of the universe. However, observing an explosion is extremely rare, so scientists run simulations.
An HPC simulation of this nature takes about four months and generates over a terabyte of visualization data. Visually analyzing the data is the next big challenge, adding weeks of labor to process and generate batch renderings.
But with NVIDIA IndeX, scientists and researchers can accelerate their visualization pipeline with real-time feedback and interactive data analysis, giving researchers more time to analyze their data and gain better insights faster.
“Analyzing a terabyte of data with batch rendering is an iterative process that takes us weeks to complete. NVIDIA IndeX allows us to interactively view the dataset and quickly focus on the interesting parts of the supernova,” says Philipp Moesta, Ph.D., of the University of California, Berkeley. “This allows us to explore more of our simulated data that is limited with batch rendering.”
NVIDIA IndeX is free for public research and universities, or you can purchase a license for commercial use.