Nvidia, with the support of Google DeepMind and Disney Research, is developing Newton, an engine that calculates physics to simulate robot movements in real conditions. This was announced by Nvidia CEO Jensen Huang at the GTC 2025 event.
Image source: nvidia.com
One of the first companies to deploy Newton will be Disney — the platform will be used to create, for example, the BDX robots based on Star Wars, one of which appeared on stage with Mr. Huang during his speech. A preliminary version of Newton with open source code will be released this year. Disney has been working on a project dedicated to the appearance of robots from the Star Wars film universe in the company’s amusement parks for several years now. Newton will be one of the technologies that will help such machines appear in Disney theme parks next year — Senior Vice President of Disney Imagineering Kyle Laughlin noted the merits of Nvidia and Google DeepMind in the implementation of the project.
Newton will help robots become more “expressive” and “learn to handle complex tasks with greater precision,” Nvidia said. The physics engine will help developers model how robots interact with objects in the outside world — one of the most complex tasks in robotics. Newton is easy to configure, Nvidia assured. It can be used to program scenarios for how machines interact with food, fabric, sand, and other deformable objects. Newton will also be compatible with Google DeepMind’s robot design ecosystem, including the MuJoCo physics engine, which simulates multi-joint movements of mechanisms.
At GTC 2025, Nvidia also unveiled the GR00T N1, a baseline AI model that will help robots better perceive and reason about their environment.