Nvidia’s New Power in A.I.

Nvidia, already the dominant supplier of the chips powering the artificial intelligence boom, is rapidly becoming something more expansive and potentially more consequential: one of the industry’s biggest financiers, dealmakers and gatekeepers.

Fresh disclosures from the company show that Nvidia now holds roughly $99 billion in equity investments, alongside another $25 billion in investment commitments, underscoring how far it has moved beyond its traditional role as a semiconductor manufacturer. And this week, that strategy took on a new dimension when Nvidia agreed to acquire Hugging Face, the influential A.I. model-sharing and developer platform, for $12.93 billion.

Together, the moves suggest that Nvidia is not simply profiting from demand for A.I. systems. It is helping shape who builds them, how they are financed and which tools developers use to create them.

From chip supplier to power broker

For the past several years, Nvidia’s position at the center of the A.I. economy has rested on its graphics processing units, or GPUs, which have become the industry standard for training and running large A.I. models. But as demand for those chips has surged, so has Nvidia’s appetite for a deeper role across the industry.

In its latest quarterly filing, covering the period ended July 26, Nvidia detailed equity holdings approaching $100 billion and large additional commitments tied to A.I. infrastructure, cloud partners and customers. The company has also said it plans to help mobilize more than $500 billion in third-party capital for A.I. infrastructure projects.

Those figures illustrate a significant evolution. Nvidia is increasingly using its balance sheet to back the broader ecosystem that buys, deploys and depends on its technology. In some cases, that support has taken the form of investments in A.I. labs, cloud providers and infrastructure companies. In others, it has involved more elaborate financing structures and customer-support guarantees tied to computing capacity.

Reuters reported in August that Nvidia had offered up to $105 billion in credit support connected to an OpenAI-linked data-center lease in Ohio, another sign of how aggressively the company is helping finance the buildout of the industry that sustains its own growth.

The strategy has obvious advantages for Nvidia. By reducing financial bottlenecks for customers and infrastructure partners, it can help speed demand for the very chips and systems it sells. But it also exposes the company more directly to the risks of an overheated market, especially if some high-profile A.I. projects struggle to turn enthusiasm into durable revenue.

The Hugging Face deal

That tension is part of what makes Nvidia’s acquisition of Hugging Face so striking.

Hugging Face is one of the most important distribution and collaboration hubs in the A.I. world, serving researchers, developers and companies that use open models and related tools. Jensen Huang, Nvidia’s chief executive, said the platform reaches about 18 million developers, 200,000 enterprise customers and 3 million models.

Clément Delangue, Hugging Face’s chief executive, said the company approached Mr. Huang during the summer, and Mr. Huang said there were other bidders. In announcing the deal, Nvidia said the acquisition would help “expand access to AI for developers and institutions worldwide.”

The purchase gives Nvidia ownership of a company that sits at a critical junction of the A.I. ecosystem: between model creators, infrastructure providers and end users. Hugging Face has become a central venue for sharing models, evaluating tools and distributing software used throughout the field. Bringing that platform under Nvidia’s control could give the chipmaker even greater sway over which models, frameworks and workloads gain momentum.

Delangue has said the founders and team will join Nvidia and that Hugging Face will continue to operate as a neutral platform within the company. Whether developers, enterprise customers and open-source communities accept that assurance may prove to be one of the most important questions raised by the acquisition.

A market growing larger — and more selective

Nvidia’s expansion comes at a moment when the economics of A.I. are entering a more complicated phase.

The frenzy around generative A.I. has driven giant capital commitments, soaring private valuations and a scramble for data centers, chips and power. But even as spending remains enormous, some investors are becoming more cautious, especially in the early-stage startup market.

Market watchers have warned that a shakeout may be coming as venture capital firms grow more selective and place greater emphasis on productivity gains, defensible products and actual customer demand rather than sheer narrative momentum. That scrutiny matters because many young A.I. companies have been able to command lofty valuations before proving that their businesses can generate lasting returns.

Nvidia’s rising role as an investor and acquirer puts it squarely inside that shift. The company has the resources to keep funding the ecosystem when traditional venture investors turn more disciplined. That could strengthen Nvidia’s influence further, particularly if startups and infrastructure builders find it harder to raise capital elsewhere.

But selectivity in funding also raises a more basic question: whether Nvidia’s growing financial exposure is generating investment returns on its own merits, or primarily helping preserve demand for Nvidia hardware in a market still fueled by extraordinary expectations.

Why this matters now

The significance of Nvidia’s latest moves lies in the breadth of control they imply.

The company already supplies the chips widely viewed as indispensable to advanced A.I. development. Now it is also financing the infrastructure needed to deploy those systems and acquiring a platform that many developers use to discover, share and build with A.I. models.

That combination could make Nvidia more deeply embedded in nearly every layer of the A.I. stack — from capital formation and data-center expansion to software distribution and developer adoption.

It also invites scrutiny. Regulators may eventually examine whether ownership of Hugging Face could affect platform neutrality or tilt the market toward Nvidia’s preferred tools and architectures. Customers and developers may ask whether a platform long associated with openness can remain meaningfully multi-cloud and multi-chip under the ownership of the industry’s most powerful hardware company.

For now, Nvidia appears to be betting that the next phase of the A.I. race will be won not only by selling the picks and shovels, but by financing the mines and owning part of the marketplace as well.