Nvidia forecasts 70% sales growth next year, signals AI spending boom has years left to run
Sign up now: Get ST's newsletters delivered to your inbox
Nvidia said it expects revenue to grow by about 70 per cent in fiscal 2028.
PHOTO: REUTERS
- Nvidia's quarterly revenue more than doubled to US$96.22 billion, with a forecast of US$108 billion for the next quarter, driven by strong AI chip demand and a key partnership with Amazon Web Services.
- Nvidia expects about 70% revenue growth by fiscal 2028 and plans to deploy 2 million additional GPUs with Amazon, but faces uncertainty in its China sales and rising competition in AI chip markets.
- Big tech is increasing AI infrastructure spending to over US$730 billion in 2026, while Nvidia's margins may face pressure from production costs and memory prices, amid efforts by rivals and customers to develop in-house chips.
AI generated
Nvidia on Aug 26 forecast a 70 per cent jump in revenue next fiscal year, underscoring unabated demand for AI computing, while warning that shortages of memory components would continue to curb how quickly it can expand.
The chip company’s shares rose nearly 5 per cent in extended trading, after having dipped over 1 per cent initially.
“AI has reached its inflection point. It’s doing useful work. Its tokens are productive and profitable. Now, compute is revenue,” Nvidia Chief Executive Jensen Huang said.
The outlook is likely to reassure investors who have questioned how long the AI spending surge can last after years of explosive growth.
By forecasting revenue growth well above Wall Street expectations and outlining demand from the biggest tech companies as well as AI labs, Nvidia is arguing that the market for AI computing is expanding rather than peaking, even as supply constraints limit how much business it can capture.
Executives shared a road map for growth over the next few years, including a ramp in its next-generation Vera Rubin processors and expanding sales at AI labs such as OpenAI.
Nvidia’s outlook for 70 per cent growth in its next fiscal year, ending January 2028, is a rare disclosure for the chip company, which typically does not issue such projections.
“We’ve never forecast or never guided to a year in advance,” Huang said.
Ahead of the results on Aug 26, analysts, on average, projected 44 per cent revenue growth in the same period.
“What makes (the forecast) even more credible is that demand is broadening beyond the original hyperscalers with AI clouds, enterprises, sovereign buyers and industrial customers now growing materially faster,” said Shay Boloor, chief market strategist at Futurum Equities.
The company said its Vera Rubin platform, which has now started shipping to customers, will account for about a fifth of its overall data center revenue in the current quarter, which ends in October.
In its fiscal second quarter ended July, data centre revenue more than doubled to US$89 billion, beating estimates of US$85.08 billion, according to LSEG data.
Supply constraints limit growth
Nvidia said it expects demand from AI labs to contribute roughly a quarter of its overall business next year, indicating a diversified customer base.
The company added that the so-called neo-clouds, which include companies like Nebius and CoreWeave, are set to exit in 2026 with more than eight gigawatts in Nvidia GPU capacity, a surge from the three gigawatts at the end of last year.
Nvidia also announced an expansion of its partnership with Amazon’s cloud computing unit, Amazon Web Services.
The pair will deploy an additional 2 million Nvidia graphics processors across Amazon’s global infrastructure in 2027 and 2028.
“We are seeing demand acceleration even at our scale. Customers’ forecasts point to our growth doubling next year. However ... we are supply-constrained,” finance chief Colette Kress told analysts on an earnings call.
Soaring memory prices and higher component costs will continue to pressure its margins, Kress said, adding that margins would bottom in the fourth quarter at roughly 71 per cent to 72 per cent, down from about 74 per cent in the third quarter.
Analysts were expecting 74.77 per cent for the third quarter.
The company forecast third-quarter revenue of US$108 billion, plus or minus 2 per cent, compared with analysts’ average estimate of US$104.19 billion, according to data compiled by LSEG.
China chip sales remain unclear
Nvidia’s China business has been highly uncertain.
In May, Washington cleared roughly 10 Chinese firms, including Alibaba, Tencent and ByteDance, to buy one of Nvidia’s most powerful AI chips, the H200, though deliveries stalled for months.
Nvidia did not include China data centre revenue in its outlook.
Nvidia began pitching its new Vera CPU to Chinese clients in June, telling them the chip could be available by August, as China separately weighed allowing top AI firms limited H200 purchases.
In July, a US Commerce Department official said shipments had begun but remained “very few”.
Nvidia’s second-quarter revenue more than doubled to US$96.22 billion, beating estimates of US$92.17 billion.
Adjusted profit was US$2.22 per share for the three months ended July 26, compared with estimates of US$2.10. REUTERS
