Nvidia forecasts 70% revenue growth as AI chip demand outstrips supply

Key points

• Nvidia shares rose nearly 6% in premarket trading on Thursday after the company issued its first-ever one-year revenue forecast.
• CFO Colette Kress said Nvidia expects 70% revenue growth for fiscal 2028, covering February 2027 to January 2028.
• CEO Jensen Huang said true demand is "much greater than 70%" but supply shortages are holding the company back.
• Nvidia's non-hyperscaler customers, including industrial and enterprise clients, delivered $40.3 billion in quarterly sales, up 138% year-over-year.
• Some analysts flagged a growing threat from custom AI chips being built by large tech firms and AI labs, including OpenAI.

Nvidia, the company that makes the specialised computer chips most AI systems run on, told investors on Wednesday that it expects its revenue to grow 70% in fiscal year 2028. The reaction was immediate. Shares climbed nearly 6% in premarket trading on Thursday.

The forecast is a first for the company. CEO Jensen Huang told investors that Nvidia has "never forecasted" a year in advance, but that it now has enough visibility into its supply chain to do so with confidence.

Why can't Nvidia just make more chips?

Demand is already running past what 70% growth would imply. Huang said demand is "much greater than 70%," but the company simply cannot produce enough product to meet it.

Two supply bottlenecks are squeezing output. TSMC, the Taiwanese manufacturer that makes most of Nvidia's chips, is stretched to capacity. Memory chips, a key component inside Nvidia's AI systems, are also in short supply.

This matters for anyone whose company is waiting to buy AI computing power. Longer waits and higher prices follow when supply is tight.

Is Nvidia's dominance under threat?

For now, no serious rival has displaced it. But analysts, as reported by CNBC Tech, did flag a "threat" on Thursday: custom semiconductors being designed in-house by the big cloud companies and AI labs, including OpenAI. These firms would rather not depend on a single supplier, and they have the money to build alternatives.

Nvidia is countering that concern by showing its customer base is broadening. A year ago, a single AI lab was responsible for a large share of demand. Today, Huang described a "golden age of new AI labs and startups, multiple frontier labs scaling in parallel" alongside industrial clients and physical AI applications such as robotics.

Non-hyperscaler customers, a group Nvidia labels ACIE (AI Clouds, Industrial, and Enterprise), brought in $40.3 billion in the latest quarter, up 138% compared with the same period a year earlier.

What does this mean for ordinary investors?

Some analysts are bullish. Siddy Jobe, senior portfolio manager at Econopolis Wealth Management, said the earnings results "tell you that the valuation today is cheap" and sees "plenty of upside." Paul Meeks, head of technology research at Freedom Capital Markets, said he does not expect a meaningful slowdown "until 2028 at the earliest."

Still, the nervousness that briefly wiped $1 trillion from chip stocks in July has not fully disappeared. Investors remain watchful over how much big tech companies are spending on AI, and whether those investments will eventually pay off.

Huang's message was blunt: AI has "reached its inflection point," the moment when a technology tips from early adoption to broad, fast-moving use. For shoppers, employees, and patients, that means AI tools are reaching more industries faster than at any point before.