Figure has partnered with Nscale to deploy up to 100,000 NVIDIA Vera Rubin GPUs for training its humanoid robot AI. The first systems are planned for the second half of 2027 in Barstow, Texas, with an initial compute commitment of $3.5 billion that could grow beyond $6 billion.
Figure and Nscale will use the NVIDIA Vera Rubin platform to provide the computing power needed to train Figure’s Helix AI models. Nscale will also make a strategic investment in Figure as part of the agreement.
Figure says its Helix system is increasingly limited by the amount of data and computing power available for training. The company recently introduced Index, a project designed to build a large humanoid robot training dataset that generates about 35 minutes of data every second.
READ ALSO: https://modernmechanics24.com/post/mit-neutrino-laser-impossible/
More data also requires more computing power to process and learn from it. Nscale’s AI cloud will provide Figure with access to the large GPU infrastructure needed to train future models for general-purpose robotics.
The planned system will use NVIDIA Vera Rubin GPUs through Nscale’s cloud platform. Figure also plans to use NVIDIA Isaac Sim, a simulation platform that can help train and test robot systems before they are deployed on physical machines.
The partnership could support the development of more capable humanoid robots. Figure says the additional compute will help it move toward its goal of putting humanoid robots into homes, while the companies will also explore using humanoids within Nscale’s supply chain.
The rollout is still several years away, with initial deployment targeted for the second half of 2027. The final scale of the deployment and the additional investment beyond the initial commitment will depend on how the partnership develops.
WATCH ALSO: https://modernmechanics24.com/post/us-coast-guard-antarctica-ice-mission/
The deal highlights the growing link between large-scale AI computing and robotics. As companies collect more physical-world training data, access to powerful GPU systems could become an important factor in developing robots that can perform a wider range of tasks.
One Response