Nvidia is moving deeper into robotics with Project GR00T, a new AI platform aimed at humanoid robots. The announcement places the chipmaker alongside many of the most visible companies building human-shaped machines, while also extending Nvidia’s robotics work through Isaac, Jetson and new programs for robotic arms and autonomous mobile robots.
Why GR00T matters
Nvidia describes Project GR00T as “a general-purpose foundation model for humanoid robots.” The goal is not a single robot, but an AI platform for companies working on the humanoid form factor.
The timing is important because humanoid robotics has become one of the most contested areas in robotics. The category has attracted large amounts of venture capital, while also drawing serious skepticism about how soon these systems can become broadly useful.
Nvidia CEO Jensen Huang framed the challenge directly: “Building foundation models for general humanoid robots is one of the most exciting problems to solve in AI today.” That statement captures why Nvidia wants to be involved. The company has already become central to AI hardware, and it has been making a broader case for itself in robotics through initiatives such as Isaac and Jetson.
GR00T is Nvidia’s way of planting a flag in humanoid robotics. It gives the company a platform story, not only a chip story, as builders search for computing, simulation, machine learning and deployment tools.
The humanoid companies in Nvidia’s orbit
The platform is being positioned for a long list of humanoid robot makers. Nvidia names 1X Technologies, Agility Robotics, Apptronik, Boston Dynamics, Figure AI, Fourier Intelligence, Sanctuary AI, Unitree Robotics and XPENG Robotics among the companies connected to the effort.
That list covers many prominent companies in the humanoid robot market. The source also notes a few notable exceptions, including Tesla.
Agility Robotics receives particular attention in the announcement. Co-founder and Chief Robotics Officer Jonathan Hurst said: “We are at an inflection point in history, with human-centric robots like Digit poised to change labor forever. Modern AI will accelerate development, paving the way for robots like Digit to help people in all aspects of daily life. We’re excited to partner with NVIDIA to invest in the computing, simulation tools, machine learning environments and other necessary infrastructure to enable the dream of robots being a part of daily life.”
Sanctuary AI co-founder and CEO Geordie Rose also emphasized the need for shared infrastructure and long-term partnerships. He said: “Embodied AI will not only help address some of humanity’s biggest challenges, but also create innovations which are currently beyond our reach or imagination. Technology this important shouldn’t be built in silos, which is why we prioritize long-term partners like NVIDIA.”
Both comments point to the same underlying issue: humanoid robots are not only mechanical systems. They need computing, AI models, simulation environments and developer tools that can work together.
Jetson Thor gives GR00T new hardware
GR00T is also tied to new Nvidia hardware. Jetson Thor is a new computer built for the humanoid form factor, including simulation workflows and generative AI models.
Nvidia says the system-on-chip includes a next-generation GPU based on NVIDIA Blackwell architecture. It also includes a transformer engine delivering 800 teraflops of 8-bit floating point AI performance to run multimodal generative AI models like GR00T.
The hardware package also includes an integrated functional safety processor, a high-performance CPU cluster and 100GB of ethernet bandwidth. According to Nvidia, this simplifies design and integration efforts.
The hardware piece matters because humanoid robotics depends on more than model announcements. Developers need systems that can run the relevant workloads and fit into robots that are expected to perceive, plan and act in physical environments.
Even with that progress, the source cautions that general purpose robots are still years off. The more immediate value may be in giving third-party developers broader access to tools that help close the gap.
Beyond humanoids: arms and mobile robots
Nvidia’s GTC robotics announcements were not limited to GR00T. The company also introduced Isaac Manipulator and Isaac Perceptor, extending its robotics push into robotic arms and autonomous mobile robotics.
Isaac Manipulator focuses on dexterity and AI capabilities for robotic arms. Nvidia describes it as offering “state-of-the-art dexterity and modular AI capabilities for robotic arms, with a robust collection of foundation models and GPU accelerated libraries.”
The company also says Isaac Manipulator provides up to an 80x speedup in path planning and uses zero shot perception to increase efficiency and throughput. The stated aim is to help developers automate a greater number of new robotic tasks.
Nvidia says it already has partners for Isaac Manipulator, including Franka Robotics, PickNik Robotics, READY Robotics, Solomon, Universal Robots and Yaskawa.
Isaac Perceptor is aimed at autonomous mobile robotics, or AMRs. It builds on Nvidia’s focus on vision processing for robotics and targets “multi-camera, 3D surround-vision capabilities.” ArcBest, BYD and KION Group have signed up.
The race Nvidia wants to supply
The broader picture is clear: Nvidia is positioning itself across the robotics stack. GR00T targets humanoid robots, Jetson Thor supports the hardware side, Isaac Manipulator focuses on robotic arms, and Isaac Perceptor addresses autonomous mobile robots.
This puts Nvidia in the middle of a developing race between humanoids and mobile manipulators. Humanoids draw attention because of their human-centric design. Mobile manipulators and robotic arms remain important because manipulation has been a core robotics challenge for decades, especially in industrial settings such as automotive manufacturing.
The next generation of systems is expected to be more dexterous and more mobile. Nvidia’s announcements suggest the company wants its platforms, chips and developer tools to be part of that shift, whichever robot form factor gains the most traction.
For now, GR00T is less about proving that general purpose humanoid robots have arrived and more about giving the field a common AI and computing foundation. In a market filled with ambition and skepticism, Nvidia’s entry is a signal that humanoid robotics is becoming a platform contest as much as a robot-building contest.