Figure has signed a partnership with cloud infrastructure provider Nscale to deploy up to 100,000 GPUs on the NVIDIA Vera Rubin platform, with an initial commitment of $3.5 billion in compute and intent to scale beyond $6 billion1. Initial deployment is targeted for the second half of 2027 in Barstow, Texas.
As part of the agreement, Nscale is making a strategic investment in Figure, and the two companies will explore scaling Nscale's supply chain with humanoid robots.
The compute is earmarked for training Helix, Figure's AI model for humanoid robotics. Figure said it is entering a phase where progress is "largely bound by data and compute needed to train Helix". The company framed the Nscale deal as providing "the compute runway needed to train the next generation of AI models to solve general robotics".
The announcement follows Figure's reveal of Index, described as the company's effort to build the most diverse humanoid training dataset ever assembled, which Figure said is generating 35 minutes of data every second.
"Our AI model, Helix, becomes more capable the same way every learned system does: with more data and compute. Today we're excited to partner with Nscale to bring the compute required to make this vision a reality," said Brett Adcock, Founder and CEO of Figure.
The source excerpt also includes a partial quote attributed to Jensen Huang, identified as founder and CEO of NVIDIA, beginning: "Humanoid robots extend physical AI into the world designed for people—opening a major new industry". The full quote is not available in the evidence.
ANALYSIS The deal's structure, pairing a multi-billion-dollar compute commitment with a strategic equity investment and a humanoid supply-chain exploration, ties Nscale's infrastructure business directly to Figure's hardware roadmap, giving both parties a stake in the other's growth. At up to 100,000 GPUs on NVIDIA's Vera Rubin platform, the cluster would rank among the larger single-customer GPU deployments disclosed to date in the robotics sector, though direct comparisons depend on final deployment scale. By publicly naming data and compute as its binding constraints, Figure is positioning Helix's trajectory as a scaling problem rather than an architectural one, a framing that justifies the capital outlay to investors and partners.