Kitware Integrates NVIDIA IndeX, Enables Interactive Volume Visualization of Large Datasets

June 20, 2016

The computational power of supercomputers has enabled researchers to run larger and larger simulations resulting in ever-growing datasets that have to be analyzed. Visualizing this data and getting real-time feedback is the best way to analyze these datasets.

NVIDIA IndeX compliments ParaView’s volume visualization capabilities and delivers interactive frame rates to scientific users analyzing large datasets.

This week in NVIDIA’s booth at ISC 2016, we are showing ParaView integrated with NVIDIA’s IndeX. IndeX is a scalable GPU-accelerated solution that delivers interactive, high-quality visualization of large volumetric data.

Interactivity is important to gather insights when one is visually analyzing data. Today, methods such as isosurfacing and 2D-slicing are commonly used to explore large datasets. While these methods provide high interactive frame rates, they encode considerably less information. Understanding spatial relationships is very important for datasets in domains such as Oil & Gas, Weather, and Computational Fluid Dynamics.

Analyzing volume data with isosurfaces is an iterative and time consuming process. One first guesses an isovalue, then applies the filter, analyzes results, and repeats the process until understanding is achieved.

A truly interactive scalable volume rendering tool allows users to visualize the entire dataset so that they can quickly identify regions of interest and focus on understanding their data. This process is significantly enhanced with the NVIDIA IndeX plug-in for ParaView.

NVIDIA’s IndeX team has worked closely with Kitware to develop a plugin that allows ParaView users access to IndeX through the familiar volume rendering interface offered by ParaView. Users can volume render larger datasets in real time, integrated with analysis results from ParaView for faster insights.

IndeX
High-quality volume visualization of large thunderstorm data with NVIDIA IndeX for ParaView plug-in. Data courtesy Leigh Orf (University of Wisconsin-Madison) and Rob Sisneros (NCSA).

The IndeX for ParaView plug-in delivers real-time performance for large datasets by distributing the data across multiple nodes within a GPU-accelerated cluster.

Since IndeX is integrated inside ParaView, users do not have to learn a new tool or change their existing workflow. Simply select the “NVIDIA IndeX” representation and use ParaView as you would normally.

Leave a Reply