Nvidia Chief Executive Jensen Huang announced on Thursday that the company anticipates doubling its chip sales volume in the coming year. The forecast underscores the semiconductor giant’s continued rapid expansion over the next six quarters.
The projection follows recent guidance from the company, which reported expectations of 70% growth for its fiscal year ending in January 2028, a period projected to generate approximately $673 billion in revenue. This outlook prompted analysts to significantly raise their expectations for the business.
Speaking to the media during a summit attended by U.K. King Charles III in Scotland, Huang explained the driving force behind this growth. “I expect Nvidia to sell twice as many chips as this next year as we do this year,” he said. He attributed the surge to the substantial economic benefits AI is delivering across various industries and economies, noting that investment demand is evident in nearly every country where the company operates.
Nvidia does not release total shipment figures for its entire product line. However, the company is best known for its data center graphics processing units (GPUs), including its Blackwell and Rubin architectures. Last fall, Huang revealed that Nvidia had shipped 6 million Blackwell GPUs within a four-quarter span.
Beyond GPUs, the corporation manufactures a diverse array of semiconductors. These include central processing units (CPUs), switch chips, components for optical networking and laptops, Jetson chips designed for robotics and automotive applications, and the processor powering Nintendo’s Switch 2 console.
Huang participated in the summit alongside representatives from major AI firms such as Google DeepMind, OpenAI, and Anthropic. The gathering focused on AI safety, a subject Huang has addressed frequently in recent weeks.
Debate regarding existential risks posed by advanced AI has intensified following warnings from researchers that humanity may no longer be able to control the most sophisticated systems. Addressing these concerns, Huang stated, “When a product is not safe, we should hold it back and keep engineering it.”
Safety first, he says. Hope they actually mean it when those models get too complex to control. The risks are real.
Nintendo using their chips is interesting, but let us be honest: the data center GPUs are where all the money is right now.
I wonder how supply chain constraints will hold up against this demand. It feels ambitious, perhaps even unrealistic for next year.
Double the chips? That is a staggering projection. The AI infrastructure build-out shows no signs of slowing down anytime soon.