Nvidia is aiming to give robotics engineers a boost in the form of a new “brain” for their robots.
Launched on Monday, the chip giant said that with the new Jetson Thor modules, robots around the world “are about to get a lot smarter.”
Nvidia launched Thor via a video (top) featuring the company’s long-time CEO Jensen Huang handing a gift box to a humanoid robot with a message card that says, “Enjoy your new brain.” The clip then shows multiple humanoid robots opening the box to find the new device that’s designed to make it a lot cleverer.
Thor, which is aimed at all kinds of robots and not solely humanoid ones, offers an impressive 7.5 times more AI computing power than its predecessor, Orin, which launched in early 2023. It also brings 3.1 times more CPU performance and comes with twice the memory, Nvidia said.
Thor is purpose-built for real-time robotics, handling high-bandwidth multi-sensor inputs with ultra-low latency. It’s also designed to run large generative AI and transformer-based models locally, reducing reliance on the cloud.
Nvidia notes how two prominent robotics companies will be using Thor in their work. U.S.-based Boston Dynamics, for example, is integrating Thor into its Atlas humanoid robot, enabling real-time processing of complex multimodal AI models to enhance its perception, decision-making, and whole-body manipulation capabilities.
Another U.S. firm, Agility Robotics, is adopting Thor for its 6th-generation Digit humanoid robot to give it stronger real-time perception and improved abilities in warehouse settings.
“The powerful edge processing offered by Jetson Thor will take Digit to the next level — enhancing its real-time responsiveness and expanding its abilities to a broader, more complex set of skills,” said Peggy Johnson, CEO of Agility Robotics. “With Jetson Thor, we can deliver the latest physical AI advancements to optimize operations across our customers’ warehouses and factories.”
Tech specialists at the likes of Carnegie Mellon, Stanford, and the University of Zurich will also deploy Thor for advanced perception, navigation, robot fleets, and sensor fusion.
Other companies using Jetson chips include Amazon, Meta, Figure, and Hexagon, while additional firms such as ChatGPT-creator OpenAI are reported to be evaluating the chips for their own use.
Robotics currently represents a mere 1% of Nvidia’s total revenue, with Huang citing the sector as its biggest growth opportunity in the coming years.
Nvidia is offering a Jetson AGX Thor developer kit that’s available now starting at $3,499. Nvidia Jetson T5000 modules start at $2,999 for 1,000 units and can be purchased from authorized Nvidia partners.
The tech giant also announced the DRIVE AGX Thor developer kit, which provides a platform for developing autonomous vehicles and mobility solutions. It’s available for preorder, with the first deliveries targeted for September 2025.
