The drive toward training humanoids at scale creates a new demand curve for the chipmaker because these models require massive data center capacity before specialized chips can be integrated into the physical robots. This strategy allows Nvidia to leverage its infrastructure leadership to support robotic intelligence training outside of traditional large language models.
Tuesday, October 20, 2026
Story
Nvidia Discusses Another $1B Investment in Figure
AI14 days ago
The robotics startup Figure plans to deploy about 100,000 Vera Rubin GPUs through Nscale starting in 2027 to train its humanoid models. This training operation will require approximately 230 MW of GPU power, according to The Information. Nvidia has held talks to invest an additional $1 billion in the firm to expand its footprint in the emerging sector of Physical AI.
Looking for more context or analysis?
Tilt chat provides reliable, tailored insights, alerts, and automations.