NVIDIA IndeX Plugin for ParaView 5.12: Feature Updates and New Licensing

May 1, 2024
Screenshot of ParaView showing large data set.

Learn how to leverage the power of GPUs with NVIDIA IndeX for interactive visualization of large-scale data sets in ParaView.

NVIDIA IndeX is a software library that enables interactive visualization and exploration of large-scale volumetric data sets on NVIDIA GPUs. It uses a distributed, out-of-core, and ray-tracing-based approach to render massive data sets with high fidelity and low latency. NVIDIA IndeX can handle data sets that are larger than the available GPU memory, and it can scale across multiple GPUs on a single node up to big clusters of many GPU nodes.

NVIDIA IndeX is included in ParaView as a free plugin, which allows to seamlessly switch between the standard ParaView rendering engine and the NVIDIA IndeX engine for volume representations. NVIDIA GPUs are required, making it possible to achieve interactive frame rates for compute and rendering. This enables smooth camera movements, live edits of the transfer function, real-time 2D slicing with multiple slices as well as interactive cropping of the data. The plugin provides a user interface for configuring the NVIDIA IndeX settings, such as the rendering quality and advanced volume shading capabilities. It supports both local and remote rendering modes of ParaView, with the option to use a thin ParaView client without a GPU (for example on Apple Mac) and rendering on remote GPU-equipped machines only.

New Features in NVIDIA IndeX for ParaView

The latest version of NVIDIA IndeX for ParaView introduces several new features and improvements, such as:

  • Based on latest CUDA 12
  • Support for latest NVIDIA Hopper and NVIDIA Ada Lovelace architecture GPUs
  • Support for NVIDIA RTX hardware ray tracing acceleration
  • Backwards compatible down to Maxwell architecture GPUs (e.g., including V100 and A100)
  • New Support for ARM architecture
  • Speed and quality improvements for unstructured grid data

The new support for ARM architectures prepares for custom builds of ParaView with the NVIDIA IndeX plugin on ARM CPUs such as NVIDIA Grace.

Major Licensing Change for NVIDIA IndeX for ParaView

We are excited to announce that the NVIDIA IndeX plugin for ParaView now comes with a full GPU cluster license by default, in contrast to its previous single-GPU license. All users of ParaView 5.12 can test and utilize the scalability of NVIDIA IndeX for a duration of two years, without limitations on the number of GPUs used, local and remote. As a benefit, this enables large-scale volume rendering on common multi-GPU systems like NVIDIA DGX out-of-the-box. After the initial period, the plugin will automatically return to the single-GPU licensing model. Further usage of this version of the ParaView plugin is determined by the general licensing rules, as listed at https://www.nvidia.com/en-us/data-center/index-paraview-plugin/. Users who are affiliated with an academic or research institution can continue to use the full license without any fees or restrictions. A commercial license can be purchased for use in production.

NVIDIA IndeX in Omniverse

The NVIDIA Omniverse platform provides new possibilities to develop and combine 3D tools, libraries, and services for creating complete simulation workflows, taking advantage of RTX-accelerated rendering and AI. NVIDIA IndeX is also available in Omniverse to address the specific needs of scientific visualization, paired with a ParaView Connector to simplify the conversion and transfer of the data to Omniverse.

We hope that you will enjoy using NVIDIA IndeX for ParaView and discover new insights from your large-scale data sets. If you have any questions, feedback, or issues, please contact us at nvidia-index@nvidia.com. We look forward to hearing from you and seeing your amazing visualizations.

Leave a Reply