Why Nvidia’s Hugging Face deal is more than a tip

This report is from this week’s The Tech Download newsletter. See what you see? You can subscribe here.

yesterday, Nvidia Confirmed what we all already knew.

The company plans to acquire Hugging Face, a New York-based startup that has built a repository of open-source AI models and has recently gained global notoriety for being the subject of a cyberattack involving rogue OpenAI models.

The price tag was as low as $12.9 billion, making it Nvidia’s second-largest acquisition after the $20 billion it paid for the assets of chipmaker Groq.

Nvidia CEO Jensen Huang told CNBC on Thursday that a valuation of nearly $13 billion was needed to fend off other bidders.

So why did the world’s most valuable company spend so much money on Hugface?

To understand that, we first need to explain in detail what Hugging Face actually does.

The startup has become the leading platform for building, sharing, and running AI models, especially in the open source and weighting ecosystems. This means that developers can edit and self-host models.

Half of Nvidia’s business “is actually primarily driven by an open model,” Huang said. The semiconductor giant is itself a major developer of open models, with Huang telling CNBC on Thursday that the company is the largest open model contributor in the world “so far.”

Hugging Face also sells infrastructure and collaboration tools to organizations working with closed, proprietary AI, as well as robot ducks.

The range of platforms is extensive. According to an Nvidia press release, more than 18 million people use the platform to share more than 3 million models and 500,000 datasets, and it is used by more than 200,000 companies.

Gil Luria, head of technology and research at DA Davidson, said in a note that the startup has become “one of the most important pieces of real estate in the AI ​​market.”

The closest thing to Hugging Face’s business was GitHub, a software repository that was acquired by Hugging Face. microsoft In 2018, it was $7.5 billion, Luria said.

Luria said one of the takeaways from the deal was that “when you allow big tech companies to acquire critical repositories, everyone else is at a disadvantage,” adding that Nvidia’s acquisition of Hugging Face is a “defensive move.”

“For large laboratories, or even worse, googleIf it owned Hugging Face, it would be in a position to slow the adoption of open source AI in the US, thereby increasing the power of the big labs.”

In short, Hugging Face is a place for AI developers to hang out, and it’s exactly the kind of space Nvidia wants to manage.

“NVIDIA provides visibility into customer preferences and the AI ​​models they use,” Naveen Chhabra, principal analyst at Forrester, told CNBC. “They can see which models are trending, what datasets customers are downloading, and which architectures are gaining traction weeks before they make it into mainstream tech news.”

Ian Fogg, research director at CCS Insight, told CNBC that the acquisition is an extension of the company’s “strategy to encourage the flourishing of a wide range of AI models, including both proprietary frontier models and open weight models.”

The chip giant has become one of the world’s biggest strategic backers of technology companies as it builds up stakes across the AI ​​stack and helps finance GPU purchases with increasingly complex financial arrangements.

Now, the company is seeking to secure one of the world’s premier AI hubs outright, beating out other applicants in the process.

Who were they, I hear you asking? Mr. Huang did not appear on CNBC’s “Squawk Box.”

“It doesn’t matter who the other bidders are,” he told CNBC. “The only thing that matters is who wins.”

As the U.S.-China AI competition continues to intensify, scrutiny of the U.S.’s dependence on China for key components used to power data centers is growing, raising the prospect of rising costs and worsening supply chain shortages for building AI.