Jensen Huang Says Nvidia Sees the AI Boom in Early, Targets 70% Revenue Growth 

Nvidia CEO Jensen Huang is betting that the artificial intelligence boom is still in its early stages, explaining that the chipmaker’s unusually broad reach across the AI industry gives it a view of future demand that few competitors can match.

Speaking at the Goldman Sachs Communacopia + Technology conference on Thursday, Huang reiterated his expectation that Nvidia’s revenue could grow by about 70% next year, extending a record-breaking expansion that has made the company one of the biggest beneficiaries of the global AI investment cycle.

“I think we could grow 70% year over year. We’re confident about that,” Huang said.

Analysts expect Nvidia to generate roughly $400 billion in revenue in its current fiscal year. A 70% increase would put next year’s revenue at approximately $680 billion, an extraordinary level of growth for a company that has already expanded at a pace rarely seen among large technology businesses.

Huang’s confidence comes as questions intensify over how long Nvidia can maintain its dominance. Amazon, Microsoft and Google are developing their own AI chips, while AI companies including Anthropic and OpenAI are also working on custom silicon. Publicly traded Cerebras and startups such as Etched are pursuing alternatives to Nvidia’s architecture.

Huang’s response is that the market misunderstands what Nvidia is actually selling.

“Most people think Nvidia builds a chip. I mean, you need airplanes to ship what we build,” he said.

The comparison captures how Nvidia’s business has evolved from its origins in PC graphics. Its modern AI systems combine processors, networking, memory and software into massive computing platforms that require substantial power, infrastructure and logistics.

“One GPU now is not $399. It’s $8.5 million dollars,” Huang said, referring to the scale of a connected Nvidia system. He described one such system as involving 2 million parts and requiring 250,000 kilowatts, adding that Nvidia ships thousands of them.

Demand, he said, is continuing to accelerate. Orders for a system combining 36 Grace CPUs with 72 Blackwell GPUs are growing by 27% month over month.

Huang’s argument for sustained growth rests on more than current orders. He says Nvidia’s position across the AI supply chain gives it an unusually detailed picture of where computing demand is developing.

“Nvidia runs every model. Every single lab can use us,” Huang said, pointing to models from Anthropic, OpenAI and Google as well as open-weight models.

“We are a foundational platform of the AI ecosystem, foundational platform of the AI industry,” he said.

That footprint extends well beyond AI laboratories. Nvidia works with memory-chip manufacturers, original equipment manufacturers, cloud providers, so-called neoclouds, AI-native companies and data-center developers.

“We’re tracking every single gigawatt of land, power, shell around the world. Literally everything on the planet,” Huang said.

In this context, “shell” refers to the physical structure of a data center before it is equipped with computing systems.

Huang said Nvidia is effectively receiving information from across the ecosystem, giving the company visibility into new data-center projects, available power, and expected computing demand.

“I mean, just think about all my partners. How many neoclouds are reporting back to us? How many OEMs are reporting back to us? How many clouds are reporting back to us? How many AI-native companies are reporting back to us? We’re working with everybody, and so we kind of know where everything is,” he said.

That breadth is central to Huang’s argument that Nvidia can forecast growth with greater confidence than a conventional chipmaker. If AI companies expand, cloud providers build more capacity and data centers secure more electricity, Nvidia stands to benefit across several layers of the resulting infrastructure build-out.

But the same interconnectedness has raised questions about Nvidia’s investments in companies that subsequently purchase its products.

The Circular-Deal Question

Nvidia has faced scrutiny over so-called circular arrangements in which it invests in AI companies that use some of their funding to purchase Nvidia hardware. The structure has prompted comparisons with earlier technology investment cycles in which suppliers and customers became increasingly financially intertwined.

Huang dismissed the characterization with characteristic humor.

“Well, it’s not circular because we put a little bit of money in, and a lot of money comes back,” he said.

“I look at the spreadsheet, we put in $1 and $100 comes back in. Is that circular? If that is, let’s do more of that,” he added.

Behind the joke, Huang said Nvidia has a process for assessing companies before investing. He insisted that the companies must have genuine customer contracts generating revenue.

“All told, he said he’s seen $100 billion worth of such contracts,” according to the discussion, adding: “I’m not taking any risks. … I need a sure thing.”

Nvidia’s extraordinary growth is largely tied to the financial capacity of the broader AI ecosystem. AI startups and infrastructure companies are raising enormous amounts of capital, while cloud providers and other technology companies are committing heavily to data centers and computing capacity.

For Nvidia, that creates a powerful feedback loop. More AI development requires more computing; more computing requires infrastructure; and much of that infrastructure currently relies on Nvidia’s hardware and networking technology.

The question is whether that relationship can remain as strong as AI markets mature.

Competition is already expanding beyond traditional GPU rivals. Hyperscalers are developing their own chips, while AI laboratories are exploring custom hardware to gain greater control over cost and performance. At the same time, AI companies that currently spend heavily on computing could eventually become more efficient in how they use infrastructure and tokens.

That creates a longer-term risk to Huang’s thesis. Nvidia’s visibility into industry demand may give it an advantage in forecasting the next stage of the boom, but it does not guarantee that today’s infrastructure requirements will remain unchanged.

The technology industry has repeatedly demonstrated that dominant platforms can eventually be challenged when customers find economic reasons to build alternatives.

For now, however, Huang sees few signs that the AI infrastructure cycle is close to exhaustion. Nvidia’s presence across chip supply, data centers, cloud platforms and AI developers gives it exposure to almost every major source of computing demand.

That helps explain why he remains confident in another year of exceptional growth. The harder question is what happens after that. If AI companies begin prioritizing efficiency over brute-force computing, or if custom chips become more competitive, Nvidia’s advantage could face a different test.