Nvidia is projected to sell about $430 billion worth of chips this year. That’s a lot of silicon.
The projection lines up with existing market forecasts for calendar year 2026. It arrives as demand for AI accelerators, the specialized chips that power data center workloads, shows no sign of cooling.
The numbers behind the projection
Start with the most recent quarter. In its August 2026 earnings release, Nvidia reported quarterly revenue of $96.2 billion.
That figure was up 106% from the same period a year earlier.
The data center business did most of the heavy lifting, contributing $89 billion. That works out to more than nine of every ten dollars Nvidia brought in during the quarter.
For context, Nvidia reported fiscal 2026 revenue of $215.9 billion. The roughly $430 billion calendar-year chip sales projection is about twice that figure.
The company’s forward guidance is even more ambitious. Nvidia’s CFO has pointed to annual revenue growth of about 70% for fiscal 2028, which would put sales near $673 billion.
CEO Jensen Huang added his own marker. He indicated Nvidia expects to sell twice the chip volume next year compared with 2026.
From gaming cards to AI infrastructure
Nvidia built its name on graphics chips. The company has since transformed into a leading supplier of AI-accelerated computing hardware.
Its current growth runs through newer GPU platforms like Blackwell and Vera Rubin. Demand for both continues to rise, according to the research behind the projection.
The customer base is also broadening. Nvidia’s buyers now include AI labs, enterprises and sovereign entities, meaning governments building their own AI capacity.
The supply problem nobody can code around
There’s a catch, and it isn’t demand. It’s supply.
The semiconductor supply chain remains constrained. Nvidia cannot simply decide to ship more chips. It depends on manufacturing capacity that takes years to expand.
As a result, the company has forecast revenue growth below what actual market demand would support. Translation: Nvidia believes it could sell more than it can make.
It also helps explain why Huang’s volume target stands out. Doubling chip output in a year is an operational goal as much as a sales goal. It depends on capacity actually arriving on schedule.
What this means
For investors, the case for optimism is fairly direct. Revenue more than doubled year over year, guidance points to roughly 70% growth for fiscal 2028, and leadership expects chip volume to double next year.
The broader signal is about AI infrastructure spending itself. A projected $430 billion in chip sales for one company in one year points to persistent global investment in the hardware behind AI. Labs, corporations and governments are all still buying.
The things to watch are straightforward. Can Nvidia actually double unit volume? Does supply loosen enough to let revenue track demand more closely? And does the fiscal 2028 target of roughly $673 billion hold as the year unfolds?
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.