2 min

Tags in this article

, , , ,

The new capabilities are aimed at helping enterprise users and creators.

This week Nvidia announced the latest release of its Omniverse Enterprise platform. Omniverse is a dynamic collaborative platform that allows artists, engineers, creators and others to collaborate remotely on virtual worlds at scale.

The recently updated 3D graphics platform “brings increased performance, generational leaps in real-time RTX ray and path tracing, and streamlined workflows to help teams build connected 3D pipelines, and develop and operate large-scale, physically accurate, virtual 3D worlds like never before”, Nvidia described.

The company added that artists, designers, engineers and developers can benefit from various enhancements across common Omniverse use cases. This includes breaking down 3D data silos through aggregation, building custom 3D pipeline tools and generating synthetic 3D data.

AI-powered services

The new release adds support for the Nvidia Ada Lovelace architecture found in modern GeForce RTX 4090 and 4080 graphics cards and the latest datacenter-scale OVX graphics computing systems for digital twins.

The platform also supports third-generation RTX technology and DLSS 3. Nvidia claims the new release delivers up to three times more performance when powered by “the latest GPU technology”, such as NVIDIA RTX 6000 Ada Generation or the NVIDIA L40.

The update also delivers new Omniverse Connectors, layer-based live workflows, an improved user experience and customization options, including multi-view ports, editable hotkeys and lighting presets. The new Omniverse Connectors allow artists and designers to connect more software such as Adobe Substance 3D Painter, Autodesk Alias, PTC Creo and Kitware’s Paraview.

Nvidia also announced the general availability of Omniverse DeepSearch, an AI-powered service that helps dev teams sift quickly through large, untagged databases of 3D models using natural language or 2D reference images. The company says the tool makes it easier to discover the models that a team needs without having to spend hours poring over a catalogue using text searches.