Jensen Huang attends a reception for the 2025 Queen Elizabeth Prize for Engineering, at St James’ Palace in London, Brirain, Nov. 5, 2025.
Yui Mok | By way of Reuters
Nvidia CEO Jensen Huang revealed in October that his firm has $500 billion in orders, in 2025 and 2026 mixed, for its chips which are on the coronary heart of the substitute intelligence growth.
For a corporation that has seen its quarterly income develop practically 600% over the previous 4 years, Huang’s assertion was an indication that Nvidia is assured of one other 12 months of sturdy — however slowing — development for its subsequent cycle of chips, implying that the AI growth nonetheless has room to run.
“That is how a lot enterprise is on the books. Half a trillion {dollars} price up to now,” Huang mentioned on the firm’s GTC convention in Washington.
Huang included 2025 income up to now, gross sales of Nvidia’s present Blackwell graphics processing models and subsequent 12 months’s Rubin GPUs and in addition associated components like networking. After parsing by means of the small print of Huang’s remarks, analysts concluded that the assertion signaled a meaningfully greater 12 months by income in 2026 than Wall Road had beforehand anticipated.
“NVDA’s disclosures counsel clear upside to present consensus estimates,” wrote Wolfe Analysis analyst Chris Caso in a November notice. Caso estimated that Huang’s information level advised information middle gross sales that may very well be $60 billion over prior calendar 2026 estimates. He has the equal of a purchase score on the inventory.
However Nvidia inventory is buying and selling 5% underneath the place it was when Huang referred to as the corporate’s shot on Oct. 28.
It is a reflection of the continued debate amongst buyers concerning the AI growth, and whether or not a handful of huge cloud firms referred to as hyperscalers and AI labs are overspending on infrastructure.
When Nvidia experiences third-quarter earnings on Wednesday, analysts polled by FactSet predict $1.25 in earnings per share on $54.83 billion of gross sales, which might be a 56% improve on a year-over-year foundation. They’re additionally searching for steering within the January quarter of $61.88 billion, which might point out a reacceleration of development.
Nvidia does not present multiple quarter of forward-looking steering at earnings. However something Huang says concerning the firm’s gross sales backlog and outlook for calendar 2026 might be scrutinized not only for Nvidia’s outlook but in addition that of the broader tech business. Analysts polled by FactSet at the moment count on $285 billion in gross sales for Nvidia in 2026.
‘Insatiable AI urge for food’
On the Washington convention, Huang mentioned the corporate has “visibility” into that income. That is not stunning — Nvidia counts practically each multitrillion-dollar tech firm as a buyer, together with Google, Amazon, Microsoft and Meta.
Throughout October earnings, all of these firms mentioned they have been boosting their capital expenditures spending on synthetic intelligence infrastructure, which implies Nvidia chips.
Hyperscalers’ rising capex displays “insatiable AI urge for food,” wrote Oppenheimer analyst Rick Schafer in a notice earlier this month. He has a purchase score on Nvidia inventory.
Nvidia has additionally been an aggressive dealmaker in the course of the quarter, and analysts will need to hear from Huang concerning the particulars of those partnerships.
The largest deal was Nvidia agreeing to take a position as much as $10 billion in OpenAI fairness in alternate for the AI startup shopping for between 4 million and 5 million GPUs over a interval of years. Nvidia additionally agreed to take a position $5 billion in former rival Intel. That deal would see the 2 chipmakers staff as much as allow Intel chips to work higher with Nvidia GPUs.
After the October quarter ended, Nvidia took a $1 billion stake in Nokia to staff as much as combine its GPUs into mobile community {hardware} of the Finnish firm. Nvidia additionally continued investing in numerous startups.
Citi analyst Atif Malik mentioned in a November notice that the cope with OpenAI, particularly, might be an investor deal with Wednesday.
“Though issues across the mixture of debt and round financing round AI capex froth exist, we basically see AI provide beneath demand,” Malik wrote. He has the equal of a purchase score on the inventory.
Nvidia has greater than 90% of the marketplace for AI GPUs. However a few of its clients — together with Amazon with its Tranium chips, Google with its TPU chips and OpenAI with forthcoming chips made in partnership with Broadcom — have promoted their customized semiconductors, or ASICs, more and more over the previous three months.
Huang typically speaks usually about Nvidia’s views on an earnings name with analysts and will elaborate about how the corporate sees rising competitors, which buyers would welcome, in response to Citi.
All of those projections are with none China gross sales. The corporate’s Chinese language-oriented chip, referred to as the H20, was successfully restricted from being exported earlier this 12 months earlier than Huang made a cope with President Donald Trump in August to get export licenses in alternate for the federal government getting 15% of China gross sales.
However since then, Nvidia representatives have made gloomy feedback about the potential of important gross sales to China, and the corporate hasn’t introduced a successor chip to the H20, which is getting previous by AI chip requirements. Schafer, the Oppenheimer analyst, mentioned he believes China may signify an over $50 billion annual income alternative.
When requested by CNBC in late October whether or not he desires to promote present Blackwell-generation chips to China, Huang mentioned: “I hope so. However that is a call for President Trump to make.”