NVIDIA Announces Project Gr00t for Robotic AI, Apple Vision Pro Integration

Riding high on a high stock price and red-hot artificial intelligence growth, Nvidia on Friday announced several initiatives to capitalize on AI, including a partnership to build better humanoid robots, a new Blackwell computer chip and a new way to design factories. and utilities using Apple's Vision Pro spatial computing platform.

The announcements came from a two-hour keynote Monday by NVIDIA founder and CEO Jensen Huang at the annual GPU Technology Conference (GTC) in San Jose, California.

Starting with the company's foundational technology, Huang noted that AI development needs will dramatically increase the amount of data involved.

“We need to build bigger GPUs,” he said, noting that the existing Hopper line is well ahead of the launch of “very, very large GPUs.”

Named after mathematician David Blackwell, Huang's new hardware is more than just a GPU, it's a platform, but it showcases the new B200 chip at its core.

“People think we make GPUs. And we do, but GPUs don't look like they used to.”

“We're starting to do really well in the next generation of robotics,” Huang said. “The next generation of robotics could be human robotics; we now have the technology necessary to imagine fully human robotics.”

To get there, Huang announced Project GR00T, the latter term referring to the Marvel Comics character. General Robot 00 Technology. On stage, Huang introduced several AI-powered robots to the audience, such as the Jetsons Tor computing platform, which is designed to perform complex tasks and interact naturally with humans.

Huang said NVIDIA is developing a complete system for training humanoid robots that can learn from image content, including videos and virtual reality, bringing together several related departments and projects at the chipmaker.

“We worked on Nvidia Isaac Lab GR00T for training at scale and built NVIDIA Osmo, which coordinates training and reference workflows across Nvidia DGX systems for training, Nvidia OVX systems for simulation, and Nvidia IGX and Nvidia AGX systems for hardware-in-the-loop verification,” the company said.

Jetson Tor is part of NVIDIA's updates to the Isaac Robotics platform, and Huang says they're training these robots for real-world adaptations in Isaac's training gym.

“In one sense, human robotics could be simpler,” Huang said. “That's because we have so much simulation training data that we can give robots because they're built in a very similar way. Human robotics can be even more useful in our world because we've created the world to be something we can interact with and do well.”

The AI ​​robotics training virtual gym is called Omniverse, and Huang showed off a demo of what he described as “kind of crazy, but very, very close to tomorrow.”

The demo featured a robotics building described as a “digital twin of a 100,000-square-foot warehouse” with autonomous systems—humans and vehicles—interacting with each other. The environment allows for rapid development of both physical plant engineering and design, but also better coordinates the work of humans and robots within it. And it's more than theoretical.

Siemens is the world's largest industrial engineering and operations platform, Siemens is building the Industrial Metaverse, and today we're announcing that Siemens is connecting their crown-jewel Xccelerator to Nvidia Omniverse.

Apple's new Vision Pro headphones are also part of the picture.

“Omniverse Cloud streams to Vision Pro… Vision Pro connected to Omniverse portals to Miniverse – it's really amazing,” Huang announced. “And because all these CAD tools and all these different design tools are now integrated and connected to Omniverse, you can have this kind of workflow.”

The automotive industry was an obvious place to take advantage of AI innovation, Huang said.

“Everything that moves will be a robot, there's no question, it's more reliable, it's more convenient,” he explained. “One of the biggest industries will be automotive — we'll build a robotic stack from top to bottom, from computer systems in self-driving cars” to manufacturing.

Huang said Jetson Tor could be deployed in autonomous vehicles. He added that it will be installed in Mercedes and Jaguar Land Rover vehicles next year.

Nvidia's CEO said the company is building artificial intelligence platforms for other companies, including 1X Technologies, Agility Robotics, Aptronic, Boston Dynamics, Figure AI, Fourier Intelligence, Holy AI, Unitre Robotics and XPENG Robotics.

“Building foundational models for general humanoid robots is one of the most interesting problems to solve in AI today,” Huang said. “The enabling technologies are coming together to lead robotics experts around the world to artificial general robotics.”

Edited by Ryan Ozawa.