MUMBAI: The AI boom still has plenty of fuel in the tank. Nvidia has forecast revenue growth of about 70 per cent for its next fiscal year, signalling that demand for the computing infrastructure powering artificial intelligence shows little sign of slowing.
The forecast, which covers the fiscal year ending January 2028, is an unusually long-range projection from Nvidia. Analysts had been expecting revenue growth of about 44 per cent over the same period, according to a Reuters report.
The upbeat outlook comes as the chipmaker continues to benefit from heavy investment by cloud providers, technology companies, AI laboratories, enterprises and governments building their own computing infrastructure.
Nvidia’s second-quarter revenue more than doubled to $96.22 billion for the three months ended July 26, beating analysts’ average estimate of $92.17 billion. Adjusted profit came in at $2.22 per share, compared with expectations of $2.10.
The company’s data centre business was the biggest driver. Revenue from the segment more than doubled to $89 billion in the quarter, exceeding estimates of $85.08 billion.
For the current quarter, Nvidia expects revenue of $108 billion, plus or minus 2 per cent. That compares with an analyst estimate of $104.19 billion.
The numbers suggest that the enormous capital spending cycle around AI infrastructure is continuing to expand rather than approaching a peak.
Nvidia is also preparing for the next phase of its chip roadmap, with its next-generation Vera Rubin platform beginning to ship to customers.
The company expects Vera Rubin to account for around a fifth of its overall data centre revenue in the current quarter ending in October, highlighting how quickly new generations of AI hardware are being absorbed by customers.
AI laboratories are also becoming a larger part of Nvidia’s business. The company expects demand from AI labs to contribute roughly a quarter of its overall business next year, pointing to a customer base that is becoming broader than the large cloud providers that initially drove much of the AI infrastructure boom.
Companies operating so-called neo-clouds, including Nebius and CoreWeave, are also expanding their Nvidia GPU capacity rapidly. Their combined Nvidia GPU capacity is expected to exceed eight gigawatts by the end of this year, compared with three gigawatts at the end of 2025.
Nvidia is also deepening its relationship with Amazon Web Services, the cloud computing arm of Amazon.
The two companies plan to deploy an additional two million Nvidia graphics processors across Amazon’s global infrastructure in 2027 and 2028. The expansion underlines the scale of computing capacity cloud providers are preparing to add as demand for AI workloads rises.
For Nvidia, this provides a sizeable pipeline of demand beyond its existing business with major technology companies and hyperscalers.
The biggest limitation may not be demand but Nvidia’s ability to fulfil it.
The company expects shortages of memory components and higher component costs to continue putting pressure on its margins. Gross margins are expected to bottom at around 71 per cent to 72 per cent in the fourth quarter, compared with roughly 74 per cent in the third quarter.
The supply constraints mean Nvidia could potentially be selling more chips if it had greater access to the components needed to manufacture and deploy them.
Nvidia’s outlook does not include data centre revenue from China, leaving an important part of the market uncertain.
Washington cleared several Chinese companies, including Alibaba, Tencent and ByteDance, to purchase Nvidia’s H200 chips in May, although deliveries were delayed. Nvidia has also been pitching its Vera CPU to Chinese customers as Beijing weighs allowing leading AI companies limited access to H200 chips.
US officials said shipments had begun in July, but volumes remained very limited.
That uncertainty means Nvidia’s latest forecast is largely built around demand outside China, making the projected 70 per cent growth even more significant.
For now, Nvidia’s numbers offer another sign that the AI infrastructure cycle is still running hot. With cloud providers adding millions of GPUs, AI labs scaling their computing needs and newer chip platforms moving into production, the spending boom that has powered Nvidia’s rise appears to have more runway ahead.
