Nvidia just posted the kind of quarter that makes you double-check the numbers. $68.1 billion in revenue. $43 billion in net income. In a single quarter. To put that in perspective, Nvidia made more profit in three months than most companies on the planet make in a year. The AI chip giant beat expectations across the board and gave guidance that surprised even the most bullish analysts.
Here's what happened.
The Numbers: Everything Beat Expectations
- Revenue: $68.1 billion, up 73% year-over-year and 20% from last quarter
- Net Income: $43.0 billion (GAAP), up 94% from $22.1 billion a year ago
- EPS: $1.76 GAAP / $1.62 adjusted vs. $1.53 expected
- Gross Margin: 75.0% GAAP / 75.2% non-GAAP
- Operating Income: $44.3 billion, up 84% year-over-year
- Free Cash Flow: $34.9 billion for the quarter, up 125% year-over-year
- Full-Year Revenue: $215.9 billion for fiscal 2026, up 65% from the prior year
Translation: Every major financial measure came in above what Wall Street expected. The number that really matters here is free cash flow: $34.9 billion in a single quarter. That's the actual cash left over after Nvidia pays all its bills and invests in its business. It tells you this isn't just growth on paper. The money is real and it's enormous.
Translation: Quick note on "FY2026": Nvidia's fiscal year runs from February to January, so their "fiscal year 2026" actually ended on January 25, 2026. It's confusing, but lots of companies do this. When Nvidia says "Q4 FY2026," they mean the three months ending in January 2026. Just think of it as their most recent quarter.
Like these translations?
Reading the full earnings report yourself? The Ask AYO extension highlights and translates the jargon in real time — so you can read any company's press release, 10-K filing, or investor call transcript and actually understand it. Free.
Get the free extensionData Center: 91% of the Entire Business
Data center revenue hit $62.3 billion, up 75% from a year ago and 22% from last quarter. This single segment now accounts for 91% of everything Nvidia sells. That's not a typo. Nine out of every ten dollars Nvidia earns comes from selling AI chips to data centers.
The biggest buyers are the companies you already know: Amazon, Microsoft, Google, Meta, and Oracle. These "hyperscalers" are spending staggering amounts on AI infrastructure. Combined, their capital expenditure plans for the year could approach $700 billion, and a huge chunk of that spending flows directly to Nvidia.
Nvidia's latest chip architecture, Grace Blackwell, accounted for roughly two-thirds of data center revenue in Q4. The company also said its "sovereign AI" business, where national governments buy AI systems to build their own local infrastructure, more than tripled to over $30 billion for the full year.
Translation: "Hyperscalers" are the massive cloud computing companies that run the internet's infrastructure. Think Amazon Web Services, Microsoft Azure, Google Cloud. They need Nvidia's GPUs to power everything from ChatGPT to AI search to autonomous vehicles. When these companies announce they're spending hundreds of billions on AI, what they're really saying is: "We're buying a lot of Nvidia chips."
Translation: "Sovereign AI" means countries building their own AI capabilities instead of relying on American tech companies. Think of it like a country deciding to build its own power grid rather than importing electricity. Governments in the Middle East, Asia, and Europe are buying Nvidia hardware to train AI models on their own soil, in their own languages, under their own rules.
The Rest of the Business: Gaming, Pro Viz, and Cars
Nvidia isn't just AI chips, even if it sometimes feels that way. Gaming revenue came in at $3.7 billion, up 47% from a year ago. That growth was fuelled by demand for the latest GeForce RTX GPUs, though the company warned that supply constraints will limit gaming chip availability going forward.
Professional visualisation was the standout surprise: $1.32 billion in revenue, up 159% year-over-year. This segment covers GPUs used for things like 3D design, video editing, and digital twins in industries like architecture and manufacturing.
Automotive and robotics brought in $604 million, up 6% from a year ago but below expectations. This covers chips for self-driving cars and robots, and it's still a small piece of the puzzle.
Translation: "Digital twins" are virtual replicas of real-world objects or systems. Imagine a car manufacturer creating an exact digital copy of their factory floor to test changes before making them in real life. Nvidia's GPUs power the graphics and simulations needed to make these digital copies work. The 159% jump in this segment suggests more industries beyond tech are starting to use Nvidia's chips.
Margins: Hardware Company With Software Profits
Here's the number that makes Wall Street's eyes water: 75% gross margin. For a company that sells physical chips, that's almost unheard of. Most hardware companies are lucky to hit 40–50%. Nvidia is operating at margins you'd normally only see from software companies like Microsoft or Adobe.
A year ago, Nvidia's gross margin was 73%. It's gone up, not down, even as the company scaled production of its newest and most complex chips. That's significant because new products usually have lower margins at launch. Nvidia guided for roughly 75% gross margin again next quarter.
Translation: Gross margin tells you how much money Nvidia keeps after paying the cost of actually making its chips. A 75% gross margin means for every $1 of revenue, Nvidia keeps $0.75 after manufacturing costs. The remaining $0.25 goes to the factories (mainly TSMC in Taiwan) that physically produce the chips. The fact this number is so high tells you Nvidia has extraordinary pricing power: customers need these chips badly enough to pay a massive premium for them.
Guidance: Next Quarter Looks Even Bigger
Nvidia guided for $78 billion in revenue next quarter (Q1 FY2027). Wall Street was expecting $72.6 billion. That's a $5.4 billion beat on guidance alone, which would be a record quarter for many companies on its own.
The company also confirmed it shipped its first Vera Rubin chip samples to customers this week. Vera Rubin is Nvidia's next-generation platform after Blackwell, promising 10 times better performance per watt. Production shipments are expected in the second half of the year.
One notable detail: Nvidia said it's not expecting any data center revenue from China in its Q1 outlook. US export restrictions continue to block Nvidia's most powerful chips from being sold to Chinese companies, cutting off what was once a significant market.
Translation: "Guidance" is when a company tells investors what they expect their results to look like next quarter. Think of it as a forecast. When a company "beats on guidance," they're saying the next few months will be even better than analysts predicted. This matters because investors are always looking forward, not backward. Strong guidance tells the market that the growth isn't slowing down.
The Bottom Line for Investors
Nvidia delivered a record quarter by every measure. Revenue up 73%, net income nearly doubled, and the company gave guidance that blew past expectations. This is a business generating $35 billion in free cash flow per quarter while growing at a rate that most companies would struggle to match at one-tenth the size.
↑ The Bull Case
AI spending is accelerating, not slowing down. The hyperscalers are set to spend close to $700 billion on capital expenditure this year, and Nvidia captures the most valuable piece of that spending. Grace Blackwell is ramping fast, Vera Rubin is on track for later this year, and sovereign AI is opening up entirely new markets. With 75% gross margins and $78 billion guided for next quarter, Nvidia's earning power keeps compounding. Enterprise adoption of AI agents is creating a new wave of demand beyond just training large language models.
↓ The Bear Case
Nvidia is priced for perfection. Even after beating estimates, the stock traded lower the next day because investors worry that expectations have gotten ahead of reality. The company trades at around 47 times earnings, which means a lot of future growth is already baked into the price. There's also the competition question: AMD is growing its data center GPU business, and the hyperscalers themselves are designing custom AI chips (Google's TPUs, Amazon's Trainium) that could reduce their dependence on Nvidia over time. Export restrictions to China are getting tighter, closing off a massive market. And the biggest risk of all: if AI spending turns out to be a bubble and companies start cutting back, Nvidia is the most exposed company in the world.
The question isn't whether Nvidia is dominant today. It clearly is. The question is whether $68 billion quarters are the new normal, or the peak of a once-in-a-generation spending cycle.
References
- Nvidia Investor Relations — Q4 FY2026 Earnings Press Release (February 25, 2026)
- CNBC — Nvidia Reports Better-Than-Expected Q4 Results (February 25, 2026)
- Fortune — Nvidia Smashes Q4 With $68 Billion in Revenue (February 25, 2026)
Ticker: NVDA (NASDAQ) · Reported: February 25, 2026
The biggest earnings, translated. Weekly.
One email a week covering what the brands you care about actually said — in plain English. No jargon, no fluff, no spam.
Unsubscribe anytime.