Jensen Huang, the founder and chief executive of Nvidia, addressed attendees at the Goldman Sachs Communacopia + Technology conference on Thursday, laying out the reasons he believes his company's dominance in artificial intelligence hardware will sustain record-level revenue growth through the end of next year. His argument rested on a single, striking claim: his company's reach across the AI industry is so extensive that he can effectively see what is coming before it arrives.

The remarks come at a time when skeptics have been questioning how long Nvidia can maintain its extraordinary run. Competition is intensifying from multiple directions. The major cloud providers - Amazon, Microsoft, and Google - are each developing their own custom AI chips. Leading AI research organizations, including Anthropic and OpenAI, are pursuing similar in-house silicon efforts. Meanwhile, newly listed companies such as Cerebras Systems and well-funded startups like Etched are positioning themselves as credible alternatives in the AI chip market.

Redefining What a GPU Actually Is

Huang pushed back against what he described as a persistent and outdated perception of Nvidia as simply a chipmaker. The company did, in fact, invent the graphics processing unit, and for much of its early history, GPUs were consumer products aimed at improving the performance of PC gaming. That era is long gone, Huang suggested, and the scale of what Nvidia produces today bears almost no resemblance to its origins.

"Most people think Nvidia builds a chip. I mean, you need airplanes to ship what we build," he told the audience. "One GPU now is not $399. It's $8.5 million dollars. That's one GPU, all connected with NVLink, 2 million parts, right? 250,000 kilowatts. That's a GPU, and we ship thousands of them."

He also pointed to strong and accelerating demand for one specific product: a computing system that integrates 36 Grace CPUs alongside 72 of Nvidia's Blackwell-generation GPUs. That system, he said, is currently seeing month-over-month sales growth of 27%, a figure he offered as evidence of sustained and expanding demand rather than a plateauing market.

The 70% Growth Projection, Revisited

Huang used the conference appearance to reinforce a revenue forecast he had introduced the previous month, when Nvidia reported another quarter of record-breaking financial results. At that time, he indicated that the company could see revenue grow by approximately 70% in the coming fiscal year. On Thursday, he repeated that projection with equal conviction.

"I think we could grow 70% year over year. We're confident about that," he said.

Industry analysts currently project that Nvidia will close its present fiscal year with roughly $400 billion in revenue. If the 70% growth rate Huang is anticipating materializes, that would translate to approximately $680 billion in revenue for the following year - a figure that would place Nvidia among the highest-grossing technology companies in the world by a significant margin.

Embedded Across the Entire AI Ecosystem

The foundation of Huang's confidence is Nvidia's position as what he called a foundational platform for the AI industry. He argued that because Nvidia's hardware and software infrastructure underpins virtually every significant AI model being developed today, the company has an unusually clear window into where the industry is headed and how fast it is growing.

"Nvidia runs every model. Every single lab can use us," he said, noting that this includes models developed by Anthropic, OpenAI, and Google, as well as open-weight models made available to the broader research and developer community. "We are a foundational platform of the AI ecosystem, foundational platform of the AI industry."

That presence, he explained, extends from Nvidia's upstream suppliers - including memory chip manufacturers - all the way through to data center construction projects and early-stage AI startups. The company is actively monitoring global infrastructure capacity at a granular level.

"We're tracking every single gigawatt of land, power, shell around the world. Literally everything on the planet," Huang said, with the term "shell" referring to the physical structure of a data center building before it has been fitted with computing equipment.

He elaborated on just how many reporting relationships feed into Nvidia's view of the market. "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.

Questions Over Circular Investment Practices

Huang's comments about the breadth of Nvidia's relationships quickly drew questions about a practice that has attracted scrutiny in recent months: the company's habit of making investments in AI startups that subsequently become customers, purchasing Nvidia hardware. Critics have drawn comparisons between this approach and similar financial arrangements that contributed to the collapse of earlier technology infrastructure companies, most notably Lucent Technologies during the dot-com era, when vendor financing and circular revenue schemes masked underlying fragility in the business.

Huang was unapologetic and, at times, playful in his response. "Well, it's not circular because we put a little bit of money in, and a lot of money comes back," he said. He followed that with a joke: "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."

Beyond the humor, Huang offered a more substantive defense of the practice. He said that before Nvidia commits capital to any company through these arrangements, it requires evidence of genuine, revenue-generating contracts with real customers already in place. He said he has personally reviewed approximately $100 billion worth of such contracts across these investments. "I'm not taking any risks. I need a sure thing," he said.

A Dominant Position, but Not Without Limits

Even with Huang's characteristic confidence on display, the longer-term picture for Nvidia carries real uncertainties. The technology industry has a well-established pattern of disruption, in which today's dominant platforms eventually face challenges from new architectures, changing workloads, or shifts in how customers choose to build and deploy systems. No company, regardless of how embedded it becomes, has proven immune to that cycle.

Huang himself acknowledged that a significant portion of current AI infrastructure spending is being driven by AI-native startups - companies that are raising large sums of capital and channeling much of it back into compute resources for their own AI development. This dynamic has been a major tailwind for Nvidia. However, as the AI sector matures and competitive pressures increase, companies are likely to become more disciplined about how they consume infrastructure and manage the cost of running AI workloads at scale. Greater efficiency in the use of compute resources could moderate the pace of hardware demand over time.

For the moment, though, Nvidia occupies an unusually powerful position across the AI supply chain, with visibility into demand that few companies at any point in the industry's history have been able to claim. Whether that position holds through the next phase of AI development remains an open question, but based on what Huang presented on Thursday, the company is entering the near term with considerable momentum and, by its chief executive's account, a clear sense of where things are heading.