Business & Markets

Figure AI and Nscale Sign $3.5 Billion Compute Deal for 100,000 NVIDIA Vera Rubin GPUs

Figure AI and Nscale have signed a strategic partnership providing Figure with access to up to 100,000 NVIDIA Vera Rubin GPUs starting in the second half of 2027, in an initial $3.5 billion compute commitment with intent to scale beyond $6 billion, as Figure’s Helix model generates 35 minutes of training data every second through its Index dataset initiative.

By Daniel Krauss | Edited by Kseniia Klichova Published:
Figure AI and Nscale Sign $3.5 Billion Compute Deal for 100,000 NVIDIA Vera Rubin GPUs
Figure 03 humanoid robots operating in a logistics environment as Figure AI secures 100,000 NVIDIA Vera Rubin GPUs through a $3.5 billion Nscale partnership, providing the compute infrastructure to train the next generation of its Helix physical AI model. Photo: Figure AI

Figure AI and Nscale have announced a strategic partnership providing Figure with access to up to 100,000 NVIDIA Vera Rubin platform GPUs, with initial deployment targeted for the second half of 2027 in Barstow, Texas. The deal represents an initial compute commitment of $3.5 billion with stated intent to scale beyond $6 billion. As part of the agreement, Nscale is making a strategic investment in Figure, and the parties will explore scaling Nscale’s supply chain operations using humanoid robots.

The partnership is directly tied to the compute requirements of Figure’s Helix AI model, which Figure CEO Brett Adcock described as the primary constraint on the company’s progress toward general robotics.

Why This Amount of Compute

Figure has entered what Adcock describes as a phase where progress is bound by data and compute rather than hardware or deployment. The company announced its Index dataset initiative last week, designed to build the most diverse humanoid training dataset ever assembled, currently generating 35 minutes of training data every second across Figure’s deployed robot fleet and data collection operations.

That data volume requires a proportional compute commitment to train against. Scaling physical intelligence – the capacity for a robot to generalize reliably across new tasks and environments – follows the same laws observed in language AI: more data and more compute produce more capable models, and the relationship has not yet plateaued.

“Our AI model, Helix, becomes more capable the same way every learned system does: with more data and compute,” said Adcock. “Today we’re excited to partner with Nscale to bring the compute required to make this vision a reality.”

The NVIDIA Physical AI Flywheel

NVIDIA’s Jensen Huang framed the deal in terms of the complete physical AI development loop. Figure’s Helix models are trained on NVIDIA Vera Rubin GPUs through Nscale’s AI cloud, validated in NVIDIA Isaac Sim before physical deployment, and run on NVIDIA GPUs inside Figure’s deployed robots.

“Nscale and Figure have activated the robotics flywheel: training Figure’s models on NVIDIA Vera Rubin through Nscale’s AI cloud, validating them in NVIDIA Isaac Sim, and deploying them on NVIDIA GPUs in Figure’s robots,” Huang said. “This is the physical AI flywheel that will accelerate the path from models to robots in the world.”

The Vera Rubin platform – NVIDIA’s next-generation AI compute architecture, featuring 27,500 GPUs in the Japan FRONTia national physical AI infrastructure announced in July – is the most capable training infrastructure NVIDIA has announced for physical AI workloads. Figure’s access to 100,000 Vera Rubin GPUs at scale represents a compute commitment comparable to what major language AI laboratories maintain for frontier model training.

Nscale’s Strategic Position

Nscale is positioning itself at the intersection of AI infrastructure and physical AI, having previously announced its Physical AI Living Lab for European robotics startups in partnership with NVIDIA. The Figure deal extends that positioning into a long-term commitment with one of the highest-profile humanoid robot companies globally, while the humanoid supply chain exploration creates a potential operational relationship beyond capital and compute.

“We’re excited to partner with Figure as physical intelligence becomes AI’s next frontier,” said Josh Payne, CEO and Founder of Nscale. “We’ve seen incredible growth with inference and agentic AI and Figure is pushing the boundaries of AI even further.”

Disclaimer: RobotsBeat is an independent media brand owned and operated by NuvexMedia LLC, publishing news, research, and insights on artificial intelligence, emerging technologies, automation, and related industries. NuvexMedia LLC invests in and collaborates with companies across the AI, Robotics, technology, software, and digital innovation sectors. These relationships do not influence RobotsBeat's editorial coverage, and the publication maintains full editorial independence to provide accurate, timely, and objective information. © 2026 NuvexMedia LLC. All rights reserved. This content is for informational purposes only and should not be considered legal, tax, investment, financial, or other professional advice.

Business & Markets, News, Robots & Robotics

More from RobotsBeat

IFA 2026 Opens in Berlin with More Humanoid Robots Than Ever Before

IFA 2026 Opens in Berlin with More Humanoid Robots Than Ever Before

by • 4 mins read

IFA 2026 in Berlin featured the highest number of humanoid robots in the trade show's history, with PrimeBOT Q1, Booster K1, EngineAI T800, Agibot A3, and Zeroth W1…