Nvidia is broadening its robotics chip family. According to ServeTheHome, the company has announced two new mid-range boards for its Jetson Thor lineup, called the T3000 and T2000, which are expected to arrive in the first quarter of 2027.
The pitch is affordability. ServeTheHome reports that the new modules are meant to be a cheaper option for customers who feel squeezed by the high cost of memory, a component that has grown pricier across the industry. In other words, these are lower-cost siblings to Nvidia's existing higher-end Jetson Thor hardware.
The positioning is echoed elsewhere. A headline from IoT News frames the T3000 and T2000 as targeting robotics "cost and power limits" — suggesting the boards are tuned not just for a friendlier price, but also for tighter energy budgets, which matters for robots and other devices that run on constrained power. IT Brief Asia describes the expansion as adding "AI" modules to the Jetson Thor family, underscoring that these chips are built to run artificial-intelligence workloads on the robot itself rather than in a distant data center.
Jetson is Nvidia's platform for so-called edge computing: compact modules that put AI processing directly inside machines like robots, drones, and industrial systems. By offering mid-range parts alongside its premium models, Nvidia is trying to widen the range of builders who can afford to put its silicon into their products.
Why it matters: as robotics and on-device AI move from labs into real-world products, the price and power draw of the underlying chips help decide which machines actually get built — and Nvidia is moving to keep more of that market on its hardware.