At the annual Siggraph event in Colorado, Nvidia announced a suite of new AI models and services designed to expedite the development of humanoid robots. The announcement was made by Jensen Huang, Nvidia’s founder and CEO, who emphasized the company's commitment to advancing robotics technology.
“The next wave of AI is robotics, and one of the most exciting developments is humanoid robots,” said Huang. “We’re advancing the entire Nvidia robotics stack, opening access for worldwide humanoid developers and companies to use the platforms, acceleration libraries, and AI models best suited for their needs.”
Nvidia introduced updated versions of Nvidia NIM, a suite of microservices that offer pre-built containers for AI tasks. Among these, the MimicGen NIM microservice generates synthetic motion data based on recorded teleoperated data from spatial computing devices like the Apple Vision Pro. The Robocasa NIM microservice facilitates the creation of robot tasks and simulation-ready environments in OpenUSD, allowing developers to design 3D testing environments.
Nvidia also unveiled Osmo, a cloud-native managed service that enables users to orchestrate and scale complex robotics development workflows across distributed computing resources, whether on-premises or in the cloud. Osmo aims to cut deployment and development times from months to under a week, streamlining processes such as reinforcement learning for humanoid robots.
The new tools are available through the Nvidia Humanoid Robot Developer Program, which provides access to Project GR00T, a general-purpose foundation model designed to power humanoid robots, and Issac Sim, a reference application for robotics simulation built on Nvidia’s Omniverse platform. Big-name robotics players, including 1x, Boston Dynamics, Figure, Fourier, and Skild AI, have been given early access to these tools.
“Boston Dynamics and Nvidia have a long history of close collaboration to push the boundaries of what’s possible in robotics,” said Aaron Saunders, Boston Dynamics’ chief technology officer. “We’re really excited to see the fruits of this work accelerating the industry at large, and the early-access program is a fantastic way to access best-in-class technology.”
At Siggraph, Nvidia highlighted the new suite of tools as a means to reduce the expensive development costs and time associated with humanoid robots. Developers can use simulation technologies to synthetically generate training data, capturing demonstrations using devices like the Apple Vision Pro. This data is then fed into the MimicGen NIM microservice to create synthetic datasets.
The Project GR00T model can be trained on data generated by MimicGen NIM, while the Robocasa NIM generates experiences to retrain the robot model. The Osmo tool automates administrative tasks by assigning jobs to different resources, alleviating the burden on developers.
“Developing humanoid robots is extremely complex, requiring an incredible amount of real data, tediously captured from the real world,” said Alex Gu, CEO of the general-purpose robotics company Fourier. “Nvidia’s new simulation and generative AI developer tools will help bootstrap and accelerate our model development workflows.”
Nvidia’s new offerings are set to revolutionize the field of humanoid robotics by providing developers with powerful tools and resources to accelerate innovation and reduce costs. With continued collaboration between industry leaders and the integration of advanced AI technologies, the future of humanoid robots looks promising.
