Nvidia projects 70% revenue surge for next year as AI infrastructure boom continues
Nvidia has projected a remarkable 70 percent revenue increase for its next fiscal year, demonstrating that demand for artificial intelligence computing infrastructure shows no signs of slowing, even as the company warned that shortages of critical memory components will continue constraining its ability to expand production.
The forecast sent Nvidia's shares up nearly 5 percent in extended trading, reinforcing confidence in the company that briefly surpassed $3 trillion in market capitalization in June 2024, joining Apple and Microsoft among the world's most valuable corporations.
"AI has reached its inflection point. It's doing useful work. Its tokens are productive and profitable. Now, compute is revenue," said Jensen Huang, Nvidia's chief executive, who has led the company continuously since co-founding it in 1993, making him one of the technology industry's longest-serving leaders.
Rare long-term guidance signals confidence
The 70 percent growth projection for the fiscal year ending January 2028 represents an unusual disclosure for Nvidia, which typically refrains from issuing such extended forecasts. The outlook substantially exceeds Wall Street expectations, with analysts having anticipated 44 percent revenue growth for the same period.
"We've never forecast or never guided to a year in advance," Huang acknowledged during the earnings call. The bold projection suggests Nvidia executives believe the AI computing market is expanding rather than approaching saturation, despite supply chain limitations.
Nvidia's data center division, which has experienced explosive expansion from $15 billion in fiscal 2023 to over $47 billion in fiscal 2024, continues to drive growth. In the second quarter ended July, data center revenue more than doubled to $89 billion, surpassing analyst estimates of $85.08 billion.
Memory bottlenecks persist
Despite surging demand, Nvidia faces ongoing constraints from shortages of High Bandwidth Memory chips, critical components for AI processors. Manufacturers including SK Hynix, Samsung, and Micron have struggled to meet the explosive demand for these specialized memory chips that enable advanced AI computing.
"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. She indicated that rising memory prices and higher component costs will pressure margins, which are expected to decline from approximately 74 percent in the third quarter to 71-72 percent in the fourth quarter. Even at these reduced levels, Nvidia's margins remain exceptional compared to the 40-60 percent typical of most semiconductor companies, reflecting the company's pricing power in the AI chip market.
Broadening customer base
Executives outlined a diversified revenue stream extending well beyond the hyperscale cloud providers that initially drove AI chip adoption. Nvidia said it expects AI research laboratories to contribute roughly a quarter of its overall business next year. Companies like OpenAI and Anthropic have consumed billions of dollars worth of Nvidia processors since 2022 to train large language models such as GPT-4 and Claude, with individual training runs requiring thousands of interconnected chips.
"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 highlighted growth among neo-cloud providers, specialized AI infrastructure companies like CoreWeave and Nebius. These firms, which have attracted substantial capital backing with CoreWeave raising $7.5 billion in debt financing in 2024 and going public in early 2025 with a valuation exceeding $20 billion, are set to exit this year with more than 8 gigawatts in Nvidia GPU capacity, up from 3 gigawatts at the end of last year.
Major AWS expansion
Nvidia announced an expanded partnership with Amazon Web Services, the dominant cloud infrastructure provider holding approximately 31 percent of the global market. The companies will deploy an additional 2 million Nvidia graphics processors across Amazon's global infrastructure in 2027 and 2028, representing a significant commitment from a critical customer.
The company's next-generation Vera Rubin platform, which has begun shipping to customers, will account for about a fifth of overall data center revenue in the current quarter ending in October, executives said.
China business remains uncertain
Nvidia's China operations continue facing significant regulatory uncertainty. U.S. export controls introduced in October 2022 and strengthened in October 2023 restrict sales of advanced AI chips to China, limiting Nvidia's ability to sell its most powerful H100 and H200 processors to Chinese customers without special licenses.
In May, Washington cleared roughly 10 Chinese firms, including Alibaba, Tencent and ByteDance, to purchase the H200, though deliveries stalled for months. A Commerce Department official said last month that shipments had begun but remained "very few." Nvidia did not include China data center revenue in its forward-looking projections.
For the third quarter, Nvidia forecast revenue of $108 billion, plus or minus 2 percent, compared with analysts' average estimate of $104.19 billion. Second-quarter revenue more than doubled to $96.22 billion, beating estimates of $92.17 billion, while adjusted profit reached $2.22 per share for the three months ended July 26, exceeding estimates of $2.10.
