Nvidia projects 70% revenue growth and $673B in sales for fiscal 2028

Nvidia told investors on Wednesday that it expects revenue to grow 70% in fiscal 2028. Chief financial officer Colette Kress delivered the number on the second-quarter earnings call. Analysts had been modelling 44% on average, according to LSEG data cited by CNBC.

That gap matters because of the size it implies. Consensus revenue for fiscal 2027, the year ending in January, sits at $396 billion. Growing that by 70% gets Nvidia to roughly $673 billion the year after. CNBC reckons that would put the chipmaker ahead of Apple and Alphabet on Wall Street projections, and behind only Amazon among US tech companies.

The forecast landed on top of a quarter that had already beaten estimates. Nvidia reported revenue of $96.2 billion for the three months ended 26 July 2026, up 106% from a year earlier. Guidance for the current quarter is $108.0 billion, plus or minus 2%, and it assumes no data centre compute revenue from China.

Nvidia, Q2 fiscal 2027	Figure	Change from a year ago
Total revenue	$96.2bn	+106%
Data centre revenue	$89.0bn	+117%
Gross margin (GAAP)	75.0%	+2.6 pts
Diluted EPS (GAAP)	$2.46	+128%
Q3 revenue outlook	$108.0bn	plus or minus 2%

What’s capping the 70% isn’t demand. The guidance could have been even higher if not for supply constraints, CNBC reported, as Nvidia grapples with shortages in parts like memory.

Our demand is much greater than 70%.

Jensen Huang, CEO, Nvidia, on the earnings call via CNBC

Huang told analysts that supply is what lets the company commit to 70%, and that it’s working with its supply chain to push past that. He said he gave a forecast this far out because he can see next year’s compute needs now. He also wanted the partners supplying land and power to work from the same set of numbers.

Some of that demand already has a name on it. Amazon and Nvidia announced an expanded partnership on Wednesday that adds another 2 million GPUs to AWS data centres in 2027 and 2028, TechCrunch reported. That’s five months after Amazon agreed to deploy more than 1 million. Nvidia has now committed $279 billion to secure supply and manufacturing capacity, up from $119 billion a quarter earlier.

The sceptical reading is concentration. Semafor noted that hyperscalers still account for a larger share of data centre revenue than the faster-growing AI cloud startups, governments and corporate customers. Huang argues the base has broadened, pointing to a category of buyer he calls ACIE: regional AI companies, neoclouds, startups and enterprises. He said those customers were previously largely “invisible”, which is a claim nobody outside Nvidia’s order book can check.

There’s a second objection, which is that Nvidia is helping fund its own customers. Business Insider reported that the company has $18 billion committed to equity investments for the rest of this fiscal year, and held $47.9 billion in private companies as of late July. That’s more than double the $22.3 billion it held at the end of the last fiscal year. It follows the compute financing program Nvidia announced earlier this month with six major financial firms, which we covered when it guaranteed its own chips to unlock $500 billion.

Kress addressed the circular financing charge on the call. “We see it differently,” she said, arguing the investments are measured against demand strength, the business they create and the equity returns, so the risk is limited. Whether that holds is the same question underneath the AI bubble argument, and one quarter of results doesn’t settle it.

Two things are worth watching next. The first is what Vera Rubin, which Huang says is now in full production, contributes to third-quarter revenue. TechCrunch reported that investors are watching those initial sales for signs demand carries into the next hardware generation. The second is memory, because the parts shortage is the stated reason the 70% number isn’t higher.