Jensen Huang says Nvidia's bets on AI companies carry 'low risk' after a record quarter
Nvidia's CEO defended the chipmaker's growing habit of financing AI startups, hours after reporting $96.2 billion in quarterly revenue, more than double a year ago.
Key points
- Nvidia posted $96.2 billion in revenue for its fiscal second quarter of 2027, more than doubling the figure from the same period a year earlier.
- Data centre revenue, the money Nvidia earns from selling chips and systems to cloud and AI companies, rose 117% to $89 billion in the same quarter.
- Nvidia has backed a $105 billion computing campus in Ohio where OpenAI will be the main tenant.
- Nvidia and several large Wall Street banks recently announced financing arrangements worth up to $500 billion for data centre construction.
- CEO Jensen Huang projected roughly 70% revenue growth for fiscal year 2028.
Nvidia made more money last quarter than most countries spend on their entire defence budgets. Now its CEO is explaining why the company is also acting like a bank.
Speaking to CNBC's Jim Cramer on Mad Money shortly after Wednesday's earnings release, Jensen Huang pushed back on critics who say Nvidia is propping up its own sales by lending money to the very customers who buy its chips. That practice, when a company finances customers who then spend that money on its own products, is sometimes called circular financing, and it helped inflate the dot-com bubble two decades ago.
Is Nvidia doing something risky?
Huang says no, and his main argument is simple: AI companies need more capital to get started than any startup in history, and traditional lenders will not touch them yet.
"They're not investment grade," Huang told Cramer, using a finance term that means a borrower is considered reliable enough for big institutional lenders. "They don't have the track record to be able to secure capital at a low cost. And this is where Nvidia could be helpful."
In plain terms: Huang believes frontier AI labs, the small group of companies building the most powerful AI systems, cannot get cheap loans from banks on their own. So Nvidia steps in.
Huang also argued the underlying hardware is not tied to any single customer. If an AI company Nvidia backed ran into trouble, he said, the computing infrastructure could be redeployed to a different operator. "The risk is low," he said.
What deals are we actually talking about?
| Deal | Amount | Details |
|---|---|---|
| Ohio compute campus | $105 billion | OpenAI as anchor tenant |
| Wall Street data centre fund | Up to $500 billion | Arranged with major banks |
| Equity stakes | Undisclosed | Includes OpenAI and Anthropic |
| Q2 FY2027 revenue | $96.2 billion | More than doubled year-on-year |
| Data centre revenue | $89 billion | Up 117% year-on-year |
Nvidia's financial clout comes from its near-monopoly on the specialised chips, called GPUs (graphics processing units, the hardware that does the heavy number-crunching AI models require), that power large language models. A large language model is the core technology behind tools like ChatGPT and Claude. Almost every major AI company buys its chips from Nvidia.
That position generated the cash Nvidia is now recycling back into the ecosystem as loans and equity investments, equity being an ownership stake rather than a simple loan.
What does this mean for ordinary people?
For most people, the most direct consequence is that the AI tools they use at work, at home or through their phones are likely to keep expanding, because the infrastructure funding them is not drying up. Nvidia's projected 70% revenue growth for fiscal 2028 suggests the companies building that infrastructure are still spending heavily.
The broader concern, which Huang did not fully dismiss, is whether some of this spending reflects genuine demand or is artificially sustained by Nvidia's own financing. That is a question investors are clearly still weighing: despite the record quarter, Nvidia shares have gained only about 12% this year, well behind the pace of 2023 and 2024.
Huang's answer is confident. "The money we've invested is going to generate tremendous returns," he said. Whether that confidence is warranted will take years to know.
Common questions
What is circular financing and why does it worry people?
Circular financing is when a company lends money to customers who then use it to buy that company's own products, creating sales that might not have happened otherwise. Critics worry it masks weak underlying demand, the same pattern that inflated and then collapsed the dot-com bubble in the early 2000s.
Does this affect me if I use AI tools?
Not directly, for now. The financing keeps the data centres running that power the AI products you use. If the investment cycle turned negative, development could slow, but there is no sign of that in current figures.
