Nvidia Forecasts 70% Growth as AI Infrastructure Spending Accelerates
Nvidia doubled quarterly revenue to $96.2 billion and forecast 70% growth next fiscal year, arguing that AI computing demand will remain supply-constrained through early 2028.
Nvidia has given the artificial intelligence investment cycle its most ambitious demand signal yet. The chipmaker reported $96.2 billion of revenue for the quarter ended July 26, up 106% from a year earlier, and said it expects revenue to grow 70% in the fiscal year ending January 2028. That long-range forecast is unusually specific for a company whose results have become a proxy for the direction of global AI capital spending.
The scale of the latest quarter was already exceptional. Data-center revenue reached $89 billion, 117% higher than a year earlier, while adjusted earnings came to $2.22 a share. Nvidia guided to approximately $108 billion of revenue for the current quarter. Reuters reported that the fiscal 2028 growth projection was well above the roughly 44% expansion expected by Wall Street analysts before the results. Shares rose close to 5% in extended trading after initially falling.
Demand broadens beyond the largest cloud groups
Nvidia is presenting the next stage of growth as broader than another round of spending by a few hyperscale cloud providers. Management expects AI laboratories to contribute roughly one quarter of the business next year. It also said newer cloud infrastructure companies are on course to end 2026 with more than eight gigawatts of Nvidia GPU capacity, compared with three gigawatts at the end of 2025.
The company expanded its relationship with Amazon Web Services, with the two groups planning to deploy another two million Nvidia processors across Amazon infrastructure during 2027 and 2028. This matters because the AI investment thesis increasingly depends on computing demand spreading from model developers into enterprises, industrial users and national infrastructure programs. A wider customer base would reduce concentration risk, although it could introduce more credit and financing risk among younger infrastructure operators.
Nvidia's outlook is constrained by the physical supply chain. Chief Financial Officer Colette Kress said customer forecasts point to growth that could double next year, but the company cannot satisfy all of that demand. Memory shortages and higher component costs are expected to pressure gross margins, which management sees reaching a low of roughly 71% to 72% in the fourth quarter, down from about 74% in the third. The forecast therefore describes addressable demand, not a guarantee that every order can be converted into revenue on schedule.
The capital cycle is becoming more complex
The results land as the largest technology companies are expected to spend more than $730 billion on AI infrastructure in 2026, according to figures cited by Reuters. That compares with about $400 billion last year. Nvidia sits at the center of this spending cycle, but it is also becoming more involved in financing capacity and supporting ecosystem companies. Those arrangements can accelerate deployment, while making it harder for investors to distinguish independent end-user demand from activity supported by vendor, cloud or project financing.
There are other limits to the optimistic case. China remains uncertain because of US export restrictions. Rapid model efficiency gains could reduce the amount of computing required for a given task, even if the number of AI tasks continues to rise. Customers are also developing custom processors, and returns on hundreds of billions of dollars of data-center investment remain uneven and difficult to measure.
Why it matters
Nvidia's forecast is important far beyond one company's earnings. It supports semiconductor suppliers, data-center developers, power producers, private-credit providers and infrastructure funds whose plans assume several more years of intensive AI construction. It also raises the benchmark those projects must meet. A 70% revenue expansion from Nvidia's current base would require not just more capacity, but sustained economic demand for the AI services that capacity enables.
For public markets, the quarter provides strong evidence that the infrastructure buildout has not yet slowed. It does not settle whether the spending will earn acceptable long-term returns. The next test will be whether customers can turn greater computing capacity into durable revenue quickly enough to support their own capital commitments, and whether supply bottlenecks can ease without undermining the pricing power currently embedded in Nvidia's margins.