Nvidia Omniverse Platform Update Unleashes OpenUSD Integration and Generative AI Capabilities for Advanced 3D Workflows

At Siggraph 2023, Nvidia announced a major update to its Nvidia Omniverse platform, providing new foundation applications and services for developers and industrial enterprises. This update aims to optimize and enhance 3D pipelines using the OpenUSD framework and generative AI.

Source: Nvidia

The Omniverse platform, designed to facilitate connection, description, and simulation across various 3D tools and applications, has now been enhanced for the creation of virtual worlds and industrial digitalization workflows. Companies like Cesium, Convai, Move AI, SideFX Houdini, and Wonder Dynamics have become connected to Omniverse through OpenUSD.

Among the key highlights of this update are advancements to Omniverse Kit, the engine for developing OpenUSD native applications and extensions, and improvements to the Nvidia Omniverse Audio2Face foundation app, as well as spatial-computing capabilities. Omniverse Kit now includes a new central repository, the Omniverse Kit Extension Registry, which allows developers to easily access, share, and manage over 600 core Omniverse extensions. This addition facilitates the process of toggling functionality on and off, thus simplifying custom app building.

Developers new to OpenUSD and Omniverse will benefit from new developer templates and resources that reduce coding efforts. Efficiency and user experience have also been boosted with new rendering optimizations that make full use of Nvidia’s Ada Lovelace architecture enhancements in Nvidia RTX GPUs. The Omniverse RTX Renderer now fully integrates DLSS 3 technology, and a new AI denoiser enables real-time 4K path tracing of large industrial scenes. The spatial integration powered by native RTX also has been enriched, with new extended-reality (XR) developer tools that offer flexibility in experiencing 3D projects and virtual worlds.

The platform updates are also visible in Omniverse foundation applications. The Omniverse USD Composer has been upgraded to facilitate the assembly of large-scale, OpenUSD-based scenes. Omniverse Audio2Face now provides access to generative AI APIs for realistic facial animations and gestures from an audio file alone. New Omniverse Cloud APIs have been built to help developers implement and deploy OpenUSD pipelines and applications.