Wednesday, November 19th 2025

NVIDIA Delivers Record Third Quarter Fiscal 2026 Results

NVIDIA (NASDAQ: NVDA) today reported record revenue for the third quarter ended October 26, 2025, of $57.0 billion, up 22% from the previous quarter and up 62% from a year ago.

For the quarter, GAAP and non-GAAP gross margins were 73.4% and 73.6%, respectively. For the quarter, GAAP and non-GAAP earnings per diluted share were both $1.30.
"Blackwell sales are off the charts, and cloud GPUs are sold out," said Jensen Huang, founder and CEO of NVIDIA. "Compute demand keeps accelerating and compounding across training and inference—each growing exponentially. We've entered the virtuous cycle of AI. The AI ecosystem is scaling fast—with more new foundation model makers, more AI startups, across more industries, and in more countries. AI is going everywhere, doing everything, all at once."
During the first nine months of fiscal 2026, NVIDIA returned $37.0 billion to shareholders in the form of shares repurchased and cash dividends. As of the end of the third quarter, the company had $62.2 billion remaining under its share repurchase authorization.

NVIDIA will pay its next quarterly cash dividend of $0.01 per share on December 26, 2025, to all shareholders of record on December 4, 2025.

Outlook
NVIDIA's outlook for the fourth quarter of fiscal 2026 is as follows:
  • Revenue is expected to be $65.0 billion, plus or minus 2%.
  • GAAP and non-GAAP gross margins are expected to be 74.8% and 75.0%, respectively, plus or minus 50 basis points.
  • GAAP and non-GAAP operating expenses are expected to be approximately $6.7 billion and $5.0 billion, respectively.
  • GAAP and non-GAAP other income and expense are expected to be an income of approximately $500 million, excluding gains and losses from non-marketable and publicly-held equity securities.
  • GAAP and non-GAAP tax rates are expected to be 17.0%, plus or minus 1%, excluding any discrete items.
Highlights

Data Center
  • Third-quarter revenue was a record $51.2 billion, up 25% from the previous quarter and up 66% from a year ago.
  • Revealed that NVIDIA Blackwell achieved the highest performance and best overall efficiency in the SemiAnalysis InferenceMAX benchmarks, while delivering 10x throughput per megawatt compared with the previous generation.
  • Announced a strategic partnership with OpenAI to deploy at least 10 gigawatts of NVIDIA systems for OpenAI's next-generation AI infrastructure.
  • Partnered with industry leaders, including Google Cloud, Microsoft, Oracle and xAI, to build America's AI infrastructure with hundreds of thousands of NVIDIA GPUs.
  • Announced that, for the first time, Anthropic will run and scale on NVIDIA infrastructure, initially adopting 1 gigawatt of compute capacity with NVIDIA Grace Blackwell and Vera Rubin systems.
  • Announced a collaboration with Intel to jointly develop multiple generations of custom data center and PC products with NVIDIA NVLink.
  • Revealed plans to accelerate seven new supercomputers, including with Oracle to build the U.S. Department of Energy's largest AI supercomputer, Solstice, featuring 100,000 NVIDIA Blackwell GPUs, plus another system, Equinox, featuring 10,000 NVIDIA Blackwell GPUs.
  • Celebrated the first NVIDIA Blackwell wafer produced on U.S. soil at TSMC's Arizona facility, representing revitalization of U.S. manufacturing as Blackwell reached volume production.
  • Unveiled NVIDIA Rubin CPX, a new class of GPU purpose-built for massive-context processing.
  • Introduced NVIDIA NVQLink, an open system architecture for tightly coupling the extreme performance of NVIDIA GPU computing with quantum processors, which will be adopted by more than a dozen supercomputing centers globally.
  • Revealed that Arm is extending its Neoverse platform with NVIDIA NVLink Fusion to accelerate AI data center adoption.
  • Revealed that Meta, Microsoft and Oracle will boost their AI data center networks with NVIDIA Spectrum-X Ethernet networking switches.
  • Introduced NVIDIA Omniverse DSX, a comprehensive, open blueprint for designing and operating gigawatt-scale AI factories.
  • Launched NVIDIA BlueField-4, the processor for the operating system of AI factories, with industry leaders including CoreWeave, Dell Technologies, Oracle Cloud Infrastructure, Palo Alto Networks, Red Hat and VAST Data building next-generation BlueField -accelerated data center platforms.
  • Partnered with Nokia to add NVIDIA-powered AI-RAN products to Nokia's industry-leading RAN portfolio, enabling communication service providers to launch AI-native 5G-Advanced and 6G networks on NVIDIA platforms.
  • Unveiled the all-American AI-RAN stack to accelerate the path to 6G with industry-leading partners Booz Allen, Cisco, MITRE, ODC and T-Mobile.
  • Teamed with Palantir Technologies to build a first-of-its-kind integrated technology stack for operational AI.
  • Set records on the new MLPerf Inference v5.1 benchmark with NVIDIA Blackwell Ultra, and won every MLPerf Training v5.1 benchmark.
  • Revealed that NVIDIA is working with partners including CoreWeave, Microsoft and Nscale to build the U.K.'s next generation of AI infrastructure, and announced an investment of £2 billion in the U.K. market.
  • Launched the world's first Industrial AI Cloud with Deutsche Telekom to power the AI era of Germany's industrial transformation.
  • Announced that NVIDIA is working with the South Korea government and industrial leaders, including Hyundai Motor Group, Samsung Electronics, SK Group and NAVER Cloud, to expand the nation's AI infrastructure with over a quarter-million NVIDIA GPUs.
Gaming and AI PC
  • Third-quarter Gaming revenue was $4.3 billion, down 1% from the previous quarter and up 30% from a year ago.
  • Launched Borderlands 4, Battlefield 6 and ARC Raiders with NVIDIA DLSS 4 with Multi Frame Generation and NVIDIA Reflex.
  • Celebrated 25 years of GeForce with the GeForce Gamer Festival in Seoul, South Korea, a live event including the latest NVIDIA RTX announcements.
  • Unveiled an NVIDIA RTX Remix update, adding an advanced path-traced particle system that enables modders to enhance traditional effects.
  • Released an NVIDIA Blueprint for 3D object generation, RTX-optimized NVIDIA TensorRT for Windows ML and performance boosts for leading AI tools on RTX AI PCs.
Professional Visualization
  • Third-quarter revenue was $760 million, up 26% from the previous quarter and up 56% from a year ago.
  • Began shipping NVIDIA DGX Spark, the world's smallest AI supercomputer, delivering NVIDIA's AI stack in a compact form factor.
Automotive and Robotics
  • Third-quarter Automotive revenue was $592 million, up 1% from the previous quarter and up 32% from a year ago.
  • Introduced the NVIDIA DRIVE AGX Hyperion 10 autonomous vehicle development platform, a reference compute and sensor architecture designed to enable automakers and developers to build safe, scalable level 4 fleets.
  • Partnered with Uber to scale the world's largest level 4-ready mobility network starting in 2027, targeting 100,000 vehicles.
  • Revealed that NVIDIA and U.S. manufacturing and robotics leaders, including Agility Robotics, Amazon Robotics, Belden, Caterpillar, Foxconn, Figure, Lucid Motors, Skild AI, Toyota, TSMC and Wistron, are driving America's reindustrialization with physical AI.
  • Announced that leading industrial solutions providers, including PTC and Siemens, introduced new services that bring
  • NVIDIA Omniverse -powered digital twin workflows to their extensive installed base of customers.
  • Unveiled NVIDIA IGX Thor, a powerful, industrial-grade platform built to bring real-time physical AI directly to the edge.
Source: NVIDIA
Add your own comment

22 Comments on NVIDIA Delivers Record Third Quarter Fiscal 2026 Results

#1
kondamin
They should use that money to build their own memory fab
Posted on Reply
#2
john_
Nvidia: We need to buy a bank.
Posted on Reply
#3
thesmokingman
kondaminThey should use that money to build their own memory fab
Ironically, Tesla might build their own fab. TSMC and Samsung cannot provide millions of AI5/6/7 chips for them and waiting 5 plus years for incumbent players to build new fabs... is too slow.
john_Nvidia: We need to buy a bank.
And they said AI bubble lol when Jensen is printing cash like they're the Fed.
Posted on Reply
#5
kondamin
thesmokingmanIronically, Tesla might build their own fab. TSMC and Samsung cannot provide millions of AI5/6/7 chips for them and waiting 5 plus years for incumbent players to build new fabs... is too slow.
Yeah I read that too, I'm starting to think he's a bit of a, how should i put it... dreamer. cough
Posted on Reply
#6
thesmokingman
kondaminYeah I read that too, I'm starting to think he's a bit of a, how should i put it... dreamer. cough
Oh yea its nuts but crazy moonshots are his standard play.
Posted on Reply
#7
mechtech
"reported record revenue for the third quarter ended October 26, 2025, of $57.0 billion, up 22% from the previous quarter and up 62% from a year ago."

In other news TMSC plans price increase on wafers up 22% from the previous quarter........................
Posted on Reply
#8
chodaboy19
The $65 billion guidance for next quarter is insane!
Posted on Reply
#9
Hecate91
john_Nvidia: We need to buy a bank.
Nvidia could probably buy several banks while still having money left over.
And they're sort of the bank in the AI bubble of money being passed around, somehow they're managing to sell AI hardware that companies have to let sit in warehouses without power or water resources to use them.
Posted on Reply
#10
wolf
Better Than Native
Neo_MorpheusOr something like that :D
I prefer to be smiling when it happens, happiness is a choice!

Posted on Reply
#11
TheinsanegamerN
thesmokingmanAnd they said AI bubble lol when Jensen is printing cash like they're the Fed.
Companies in bubbles tend to make a lot of money on an unsustainable business practice being sustained by intangible things like hype as opposed to real world profit.

That's how bubbles work. OpenAI doesnt make a profit and is crying to the government for a bailout. No AI model has a path to profitability.
Posted on Reply
#12
Hecate91
wolfI prefer to be smiling when it happens, happiness is a choice!

Happiness? Be happy because the more you buy, the more you save! But more like cognitive dissonance from buyers at this point in both consumer and datacenter/AI, which is reaching unprecedented levels as Nvidia is controlling the market, sponsoring media, propping up an entire economy on hype and trust based on nothing but build more servers with the hopes that AI will take over everyone's jobs, more power than any one company should be allowed to have, IMHO. Great for shareholders though, of course.
Posted on Reply
#13
Bwaze
"Third-quarter Gaming revenue was $4.3 billion, down 1% from the previous quarter "

Do these numbers have any relation on how many cards they sell? "Gaming and AI PC" is now less than 8% of revenue. Remember when just Gaming was more than 50%, and it actually meant cards sold to gamers... For playing games? Era that won't come back.

And how much of that 65 billion dollars is just circle jerk money of AI companies investing in Nvidia because Nvidia invested in them? Doesn't matter, money flows!
Posted on Reply
#14
Icon Charlie
TheinsanegamerNCompanies in bubbles tend to make a lot of money on an unsustainable business practice being sustained by intangible things like hype as opposed to real world profit.

That's how bubbles work. OpenAI doesnt make a profit and is crying to the government for a bailout. No AI model has a path to profitability.
Your'e right about AI not making a profit. Also too the excessive amount of electricity to run those famrs you know.

Someone also off handedly mentioned about the software bubble. Well Bubbles can last for years. Like the DoT.com bubble which was 3.5 to 4 years (depends whom you ask). The housing bubble was also 3.5+ years.
Hecate91Happiness? Be happy because the more you buy, the more you save! But more like cognitive dissonance from buyers at this point in both consumer and datacenter/AI, which is reaching unprecedented levels as Nvidia is controlling the market, sponsoring media, propping up an entire economy on hype and trust based on nothing but build more servers with the hopes that AI will take over everyone's jobs, more power than any one company should be allowed to have, IMHO. Great for shareholders though, of course.
IMHO from what I have seen is the concept of AI is being used as an excuse to lay off tech workers/people/undesirables that corporations are correcting for the past 2 years. The AI excuse more so than last year. IMHO most of the layoffs are from the Covid era when corporations hired way more people than they really needed.

Now again from my persepctive, and keeping (mostly) politics aside, Corporations are/have been looking what Musk Did with Twitter. One of the Baby Sitting Factories during the Age of Free Money. Once he acquired Twitter, he cut 6000+ employees, saving over 1 billion dollars per year. It's leaner, It's meaner, and throws sand in the face of people in the tech industry that wanted him to fail.

Of course it did not soooo. They needed a real excuse to cut people from their baby sitting factories and what do you know..

WE GOT AI!! It will do everything and make coffee too!!! Well as we all know that AI is still whatever input data is put in. So Garbage in... Garbage out... BUT I can sure tell you that it makes one hell of an excuse to trim the fat of employees they do not need.

I kind of laugh at all of this now. 15 years ago I've seen this train crash coming. The Tech Industry could not sustain itself by holding on to a Hipster's hand and handing out participation trophies. During the age of Free Money you could get away with this. But all it did was raise the price of living and lower the quality of life for the rest of us.

Now that we are living in a post millennial world corporations (sometimes) realize their mistake and with the excuse of AI we have had hundreds of thousands of tech jobs being replace by a HAL-9000 (or so they want you to think that) and the share holders are all happy.

Heh. AI... whether you like it or not it's just another tool for the corporation overlords to use.

We may not have true robot servants yet.
But it does make HR happy.
Posted on Reply
#15
R0H1T
thesmokingmanAnd they said AI bubble lol when Jensen is printing cash like they're the Fed.
Well you probably haven't followed (too)many bubbles then, maybe try looking up 2008 (subprime) bubble or China's real estate bubble :rolleyes:

Just because they're making money hand over fist doesn't mean it's not gonna pop, in fact it reinforces that notion!
Posted on Reply
#16
Lew Zealand
Oof, the gaming GPU division can't even cover Nvidia's tax bill any longer, shriveling like a disused appendix. At least on paper the line item gives the appearance of being diversified I guess.
Posted on Reply
#17
kondamin
R0H1TWell you probably haven't followed (too)many bubbles then, maybe try looking up 2008 (subprime) bubble or China's real estate bubble :rolleyes:

Just because they're making money hand over fist doesn't mean it's not gonna pop, in fact it reinforces that notion!
Nvidia is making money, loads of it but they aren’t spending it recklessly.
When the bubble pops their valuation will collapse but they will remain a healthy company that posts profits.
Much smaller ones as the demand will shrink to normal proportions.

The rest of the industry including tsmc which made huge investments building fabs won’t be as lucky. They will live but they will feel the crunch.

the rest of the economy that is taking loans on these BS numbers and banking their futures on them staying there. Ff’d
I think I saw somewhere that a fifth of the global economy is now just 7 companies
Posted on Reply
#18
dyonoctis
kondaminThe rest of the industry including tsmc which made huge investments building fabs won’t be as lucky. They will live but they will feel the crunch.
TSMC will probably be fine, given how constrained their supply line could become before the AI craze. I still remember how Apple getting dibs on the new processes would have an impact on the release schedule of everyone else. The AI burst will probably be a stabilization of the market, rather than a total collapse.

Generative AI and LLM are getting all the headlines, but with new regulations in place, other types of models are becoming a must in some industries ( like driving assistance being mandatory in new cars in Europe )
Posted on Reply
#19
R0H1T
kondaminNvidia is making money, loads of it but they aren’t spending it recklessly.
When the bubble pops their valuation will collapse but they will remain a healthy company that posts profits.
Much smaller ones as the demand will shrink to normal proportions.
What does that have to do with their explosive revenue stream "potentially" collapsing?

I think many of you still don't understand what massive bubbles popping means for the larger (economic) landscape, now of course it could be something close to a soft(?) landing which probably won't bankrupt too many of them but the music should stop at one point.

You're also forgetting the massive layoffs which are being planned by many of these mega corps like Apple, MS, Amazon, Google, Meta etc.
kondaminThe rest of the industry including tsmc which made huge investments building fabs won’t be as lucky. They will live but they will feel the crunch.
They'll be fine, although margins will go down for sure. Consider how memory, NAND or DRAM, cycles work & it'll likely follow the same path.
kondaminthe rest of the economy that is taking loans on these BS numbers and banking their futures on them staying there. Ff’d
Yes & what happens when OpenAI can't repay one of their loans? What 2008 taught it is that when confidence in one firm/sector goes the ripple effects are felt across the board and with the impending job losses you can imagine this could turn out much much worse than that.

This is a fundamentally bigger/harder problem than 2008 one IMO, it's akin to 1929/30 in that sense especially if any of the worse case scenario plays out.
Posted on Reply
#20
Bwaze
Nah, we'll tie national security to AI, so the bubble popping won't happen - because we'll be inflating it forever, thus solving the problem once and for all!
Posted on Reply
#21
R0H1T
The insane spending on hardware in unsustainable, so yes it will pop especially when the jobs being sidelined would be lost forever.

Now would I say that this is likely the end of U-many-tee as we know it ~ probably not, but the churn will be massive over the next decade or two at least.
Posted on Reply
#22
Panther_Seraphin
kondaminThey should use that money to build their own memory fab
Problem with Fabs is they are a massive resource drain at all times and when they arent working right they are a massive noose around your neck

Only gotta look at AMD before it managed to spin off Global Foundries AND get out of their locked in contracts with them and now Intel with their floundering since 14nm.



IF they did go down the foundry route they would ideally need to invest in some Power IC development etc to vertically integrate everything on their FE/Professional LIne of cards. Good news for nVidia as they could then mandate things quite easily in terms of Voltage standards etc going forward.
Posted on Reply
Add your own comment
Nov 19th, 2026 19:54 PST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts