Nvidia has unveiled its latest artificial intelligence chip, the Jetson AGX Thor, a £2,700 “robot brain” designed to power the next generation of humanoid and industrial machines.
The chip delivers up to 7.5 times the performance of its predecessor and offers significantly improved energy efficiency. It is capable of running multiple AI models simultaneously, from natural language processing systems such as ChatGPT to advanced computer vision tools that allow robots to perceive and respond to their environment in real time. By handling these tasks locally rather than relying on distant data centres, robots can operate more smoothly, safely, and efficiently.
Chief executive Jensen Huang has described the launch as part of a shift towards what he calls “physical AI”. This is the stage where artificial intelligence moves beyond digital applications like chatbots and image generators into real-world tasks performed by robots.
Nvidia has already secured interest from leading robotics firms, including Amazon Robotics, Boston Dynamics, and Agility Robotics. While robotics chips currently account for only a small fraction of Nvidia’s revenue, the company sees significant growth opportunities in sectors such as logistics, healthcare, manufacturing, and autonomous transport.
The Jetson Thor developer kit will begin shipping next month, with Nvidia positioning it as a foundation for robots capable of responding quickly and reliably in complex, real-world environments.
