Figure Locks In 100,000 Nvidia GPUs: The $3.5B Bet That Humanoid Robots Are AI's Next Frontier


Humanoid robotics company Figure has signed a multi-year strategic partnership with AI infrastructure firm Nscale, announced September 3, 2026: an initial $3.5 billion compute commitment scaling to over $6 billion, with up to 100,000 Nvidia GPUs on the Vera Rubin platform to train Figure's Helix robot models. After chatbots, the next scaling race is "physical AI" — and this deal puts real money behind that thesis. Here's what was announced and what it signals.

The deal, in concrete terms

  • Nscale becomes Figure's preferred compute provider — and a shareholder
  • Commitment starts at $3.5B with a framework to exceed $6B
  • Up to 100,000 Nvidia GPUs; initial deployment targeted for second half of 2027 in Barstow, Texas
  • Nvidia supplies the Vera Rubin platform and Isaac Sim validation tools

Figure CEO Brett Adcock: "To bring humanoid robots to every home in the world, we are largely constrained by data and compute." Nvidia CEO Jensen Huang called it "the physical AI flywheel that will accelerate the path from models to robots."

Why robots need 100,000 GPUs

Chat models learn from text, which the internet has in abundance. Robots must learn movement and perception — grasping objects, walking, working in a kitchen — and real-world robot data is scarce and slow to collect. That pushes training toward massive simulation and video learning, which consumes compute at frontier-LLM scale. This contract confirms, in dollars, the view that the next scaling curve belongs to embodied AI.

The context: capital orbiting Nvidia

The announcement landed the same week Nvidia's roughly $12.9B acquisition of Hugging Face was confirmed, and reports point to a large Nvidia investment in Nscale's pre-IPO round as well. The pattern: the chipmaker is putting capital across the whole ecosystem — from model distribution to robot-training infrastructure. Figures beyond the official releases may shift with regulatory review; treat them as reported, not settled.

What it means for the rest of us

  • No robots in your kitchen yet. Initial deployment is late 2027, and it's training infrastructure — not a product launch.
  • A leading indicator for skills. Simulation, robotics-plus-ML, and safety validation roles are likely to grow ahead of consumer products.
  • The numbers are frameworks. "$6B" is a ceiling in a multi-year agreement, not money already spent.

Frequently asked questions

What is Helix?

Figure's family of AI models for humanoid control, integrating vision, language, and action. Detailed specifications are largely unpublished, so rely on official releases for specifics.

How unusual is this scale?

A 100,000-GPU reservation rivals the training clusters behind frontier language models. For a robotics company, it's an early first — a step-change in what physical AI investment looks like.

Is this investment advice?

No. This article is informational; the companies involved include private firms and framework-stage figures.

Related on AI Learning Lab: Nvidia Buys Hugging Face for $12.9 Billion · OpenAI Unveils GPT-6 Astra

Comments

Post a Comment

Popular posts from this blog

AI Search vs Traditional Search: How to Use Each in 2026

Image
Something changed in how people find things, and the numbers are unusually clear about it. According to a SparkToro analysis of Similarweb clickstream data published in June 2026, about 68% of U.S. Google searches between January and April 2026 ended without a click — up from roughly 60% in 2024. Answers are arriving without anyone visiting a page. That makes "which kind of search should I use?" a question worth having a real answer to. What the data actually says Two figures from that analysis matter most. AI Overviews now appear on more than 20% of Google searches, and click-through rates fall by roughly 60% when they do . The behavior shift is not that people stopped searching — it is that a search increasingly ends where it used to begin. For anyone doing research, the tradeoff is speed against provenance. You get an answer faster, and you can see less about where it came from. When AI search is the right tool You do not know the vocabulary yet — "the clause that l...
Read more

Free vs Paid AI Tools: When Is Upgrading Actually Worth It?

Image
Every AI assistant now has a free tier that's genuinely capable, and a paid tier that costs roughly the price of a couple of lunches per month. So the question isn't "is AI worth paying for" in the abstract — it's whether your specific usage crosses the line where paying changes anything . This article gives you that line. What you actually get when you pay Pricing and features shift constantly, so ignore the marketing lists and look at the three differences that persist across providers: Priority access. Free tiers get throttled or queued at busy hours. If your work happens during peak times, this is the difference between a tool you rely on and a tool that sometimes isn't there. The current top model. Free tiers usually route you to smaller or older models, with limited access to the best one. For simple tasks this doesn't matter. For reasoning-heavy work, it's the whole ballgame. Data handling. This is the one most people overlook. Paid and busin...
Read more

What Is a Small Language Model (SLM)? On-Device AI Explained

Image
Most of the attention in AI goes to models that get bigger. A quieter trend runs the other way: models small enough to run on a laptop, a phone, or a server in your own building. These are usually called small language models, or SLMs, and they are showing up inside products you already use — often without being announced. Here is what they are and when the small option is the better one. What counts as "small" There is no official threshold, but the working definition in most technical writing puts SLMs under roughly 10 billion parameters, with the common range falling between about 1 billion and 7 billion. Frequently cited examples include Microsoft's Phi-3 Mini at 3.8 billion parameters, Llama 3.2 3B, and Mistral 7B. The contrast is with frontier models, which run to hundreds of billions of parameters and, in the largest cases, over a trillion. Parameters are the adjustable values a model learns during training — loosely, a measure of how much the model can store. Fewe...
Read more