How much revenue did Nvidia report for Q1 fiscal 2027?
Nvidia reported Q1 fiscal 2027 revenue of $81.615 billion for the quarter ended April 26, 2026, according to its May 20, 2026 earnings release.
How much was Nvidia data center revenue in Q1 FY27?
Nvidia reported record Data Center revenue of $75.2 billion, up 21% from the prior quarter and 92% from a year earlier.
What was Nvidia's Q2 fiscal 2027 revenue guidance?
Nvidia guided Q2 fiscal 2027 revenue to $91.0 billion, plus or minus 2%, and said the outlook assumes no Data Center compute revenue from China.
Did Nvidia beat Wall Street expectations?
Yes. AP reported that FactSet consensus expected $78.91 billion in revenue and $1.75 in earnings per share, while Nvidia reported $81.615 billion in revenue and official non-GAAP diluted EPS of $1.87.
Why did Nvidia increase its dividend and buyback authorization?
Nvidia returned about $20 billion to shareholders in Q1, approved an additional $80 billion buyback authorization, and increased its quarterly dividend to $0.25 per share after producing $48.6 billion of free cash flow.
This article is for informational purposes only and does not constitute financial, investment, tax, or legal advice. Market data, tax rules, and prices can change after the article date. TECHi and its authors may hold positions in securities or digital assets mentioned. Always conduct your own research and consult a licensed financial, tax, or legal professional before making decisions.
Jazib Zaman founded TECHi in 2010 and is chief executive of its publisher, TechAbout LLC. He writes about AI infrastructure, semiconductors and the companies financing them, from Palantir's growth expectations to the toolmakers behind AI memory demand. He is a former member of the Forbes Technology Council.
01The quarter-Nvidia reported $81.615 billion in Q1 fiscal 2027 revenue, up 20% sequentially and 85% from a year earlier.
02The engine-Data Center revenue reached $75.2 billion, with compute at $60.4 billion and networking at $14.8 billion under the old reporting view.
03The guide-Q2 fiscal 2027 revenue guidance is $91.0 billion, plus or minus 2%, while still assuming no China Data Center compute revenue.
04The cash signal-Nvidia generated $48.6 billion of free cash flow, returned about $20 billion to shareholders, raised the dividend, and added $80 billion to repurchase capacity.
Nvidia did not merely beat earnings expectations on May 20, 2026. It forced the market to admit that the AI infrastructure cycle is no longer a story about one hot product line. It is becoming a full economic layer, with compute, networking, software, edge devices, sovereign infrastructure, industrial systems and shareholder returns all now visible in the same income statement.
That is the real story in Nvidia's Q1 fiscal 2027 results. The headline number was enormous: $81.615 billion of quarterly revenue for the period ended April 26, 2026, up 20% from Q4 fiscal 2026 and up 85% from Q1 fiscal 2026. Data Center revenue was $75.2 billion, up 21% sequentially and 92% year over year. Q2 guidance came in at $91.0 billion, plus or minus 2%, and Nvidia said it is still assuming no Data Center compute revenue from China.
That last clause matters. The quarter was not just bigger than expected. The next quarter was guided higher while leaving a major geography out of the compute assumption. If an ordinary company had printed that combination, the stock would probably have exploded after-hours. Nvidia is not an ordinary company anymore. It closed Wednesday at $223.47, and AP reported that shares dipped slightly after-hours to about $222.12, even after the beat.
That muted reaction is not irrational. It is the tax of greatness. Nvidia has become so central to the market that investors no longer ask whether it is growing. They ask whether the next evidence point is large enough to defend a company worth roughly $5.4 trillion.
The one-sentence read
Nvidia delivered one of the strongest quarters ever reported by a public technology company, but the honest read is not simply "AI demand is strong." The better read is this: Nvidia is converting more of the AI bottleneck stack into Nvidia revenue, and Wall Street is trying to decide how long that conversion can keep happening before physics, politics, budgets or competitors interrupt it.
That distinction is important. Revenue growth of 85% at this scale is almost absurd. But the market is no longer paying Nvidia for being a chip supplier. It is paying Nvidia for behaving like the toll road of AI production. Every time a model is trained, deployed, served, accelerated, networked, cooled, upgraded or moved closer to the edge, Nvidia wants a revenue line attached to that activity.
The quarter says that strategy is working. The valuation says it has to keep working.
The numbers that matter
Start with the official scoreboard. Nvidia reported $81.615 billion in revenue. GAAP gross margin was 74.9%, while official non-GAAP gross margin was 75.0%. GAAP operating income was $53.536 billion, up 147% from a year earlier. GAAP net income was $58.321 billion, up 211%. GAAP diluted EPS was $2.39.
On Nvidia's updated non-GAAP basis, which now includes stock-based compensation expense, operating income was $53.783 billion, net income was $45.548 billion and diluted EPS was $1.87. That methodology change is not trivia. Beginning in Q1 FY27, Nvidia stopped excluding stock-based compensation from non-GAAP financial measures. The cleaner read is that the company is making its adjusted numbers more conservative at the exact moment its absolute profit pool is becoming extreme.
The Q1 FY27 investor presentation shows why the market is staring at the same data from two angles. Revenue jumped from $44.1 billion in Q1 FY26 to $81.6 billion in Q1 FY27. Free cash flow rose from $26.1 billion to $48.6 billion. Non-GAAP EPS went from $0.78 to $1.87, even under the new presentation.
AP, citing FactSet, reported that analysts expected $78.91 billion in revenue and $1.75 in earnings per share. Nvidia beat the revenue bar by roughly $2.7 billion and beat the EPS bar on the official non-GAAP number. But the bigger number was the guide. The market expected something in the high $80 billions for the next quarter; Nvidia guided to $91 billion.
The core data is almost unbelievable:
Revenue was $81.615 billion, up 85% from a year earlier.
Data Center revenue was $75.2 billion, up 92% from a year earlier.
GAAP gross margin was 74.9%, up 14.4 percentage points from the prior-year quarter.
Official non-GAAP diluted EPS was $1.87, up 140% from a year earlier.
Free cash flow was $48.6 billion, up from $26.1 billion a year earlier.
Q2 fiscal 2027 revenue guidance was $91.0 billion, plus or minus 2%.
That is not a normal beat. That is a market structure update.
Data Center is no longer one business
The most important accounting change is easy to miss. Nvidia is changing its reporting framework. Instead of leaning on the old buckets in the same way, the company is now splitting around two market platforms: Data Center and Edge Computing.
Inside Data Center, Nvidia will report Hyperscale and ACIE. Hyperscale means public clouds and the largest consumer internet companies. ACIE means AI Clouds, Industrial and Enterprise. In plain English, Nvidia is separating the cloud giants from the broader universe of AI factories, sovereign projects, industry-specific clusters and enterprise systems.
That is smart reporting, but it is also a strategic message. Nvidia wants investors to stop thinking about AI demand as "Microsoft, Amazon, Google and Meta bought more GPUs." Those buyers are still central, but the company is making room for a second demand pool: governments, industrial companies, AI cloud specialists, model builders and enterprises that want purpose-built AI capacity.
The investor presentation put numbers behind it. Hyperscale revenue was $37.9 billion. ACIE was $37.4 billion. That symmetry is the overlooked story. Nvidia is telling investors that the non-hyperscaler AI factory market is already roughly the same size as the hyperscaler bucket in this reporting frame.
Under the previous sub-market view, Data Center compute revenue was $60.4 billion, up 77% year over year and 18% sequentially. Data Center networking revenue was $14.8 billion, up 199% year over year and 35% sequentially. Networking is not a side dish anymore. In an AI factory, GPUs do not matter if the cluster cannot move data, synchronize work and keep utilization high. The faster networking grows, the more Nvidia looks like an infrastructure platform rather than a silicon vendor.
This is one of the reasons the bear case keeps slipping. Bears keep looking for a GPU cycle. Nvidia keeps trying to turn the cycle into an operating system for AI infrastructure.
The China exclusion makes the guide more powerful
The most important risk line in the release is also the most bullish line if demand holds: Nvidia is not assuming Data Center compute revenue from China in the Q2 outlook.
A $91 billion revenue guide would be impressive in any case. It is more impressive when it excludes a major compute market. It suggests the rest of the world is absorbing supply at a level that can carry the business even with China restricted.
But this is where honest commentary has to cut both ways. Excluding China from guidance is not only a sign of strength. It is a reminder that Nvidia's addressable market is being shaped by governments as much as customers. Export controls can protect national security goals, but they also encourage local alternatives, change buying behavior and make long-term market share less predictable.
Nvidia can beat numbers without China. That does not mean China is irrelevant. It means the company is so supply constrained, and non-China demand is so deep, that the near-term income statement can absorb the gap. The longer-term question is whether China restrictions create a parallel AI hardware ecosystem that becomes harder to recover later.
The stock market often treats geopolitical risk as a headline. In Nvidia's case, it is embedded in revenue architecture.
The cash machine changed the shareholder story
The quarter also changed the capital return conversation. Nvidia returned about $20 billion to shareholders through repurchases and dividends during Q1. It ended the quarter with $38.5 billion remaining under its share repurchase authorization. Then, on May 18, 2026, the board approved an additional $80 billion buyback authorization.
The company also raised its quarterly dividend from $0.01 per share to $0.25 per share, payable June 26, 2026 to shareholders of record on June 4.
That dividend increase is not the main event. Nvidia is not becoming a classic income stock. The dividend is more like a signal flare: the company is producing so much cash that the old token dividend looked silly. The buyback is the bigger tool, but even there, investors should be careful. An $80 billion authorization is huge in absolute terms. Against a $5 trillion-plus market value, it is meaningful but not transformational.
The more important point is optionality. Nvidia can fund aggressive R&D, support an enormous supply chain, invest strategically, return tens of billions to shareholders and still guide growth higher. That is why the company has become difficult to value using old semiconductor mental models.
A semiconductor downcycle company hoards cash because the next downturn can be brutal. Nvidia is behaving like an infrastructure monopoly that happens to manufacture through partners. That does not remove cyclicality, but it changes the debate.
Why the stock did not rip higher
A slight after-hours dip after this print may look strange, but it is not difficult to explain.
First, the market expected greatness. The old Nvidia earnings game was simple: analysts modeled a number, Nvidia beat it, the stock rose. That pattern gets harder when the company becomes the largest weight in the AI trade. The upside surprise has to beat not only published estimates but also private expectations, options positioning and the market's mental model of hyperscaler capex.
Second, the Q2 guide is massive, but investors know the next question starts immediately. If Q2 is $91 billion, what is Q3? What is the exit rate into Rubin? How much of the growth is supply release versus durable end-demand? What happens when customers finish the first phase of AI factory buildouts and begin asking about returns on invested capital?
Third, Nvidia's valuation compresses emotion. At this scale, a great quarter can be correct and the stock can still pause. That does not make the quarter weak. It means the market is sorting out whether the next leg of the story is upside revision or expectation saturation.
This is the hardest part of covering Nvidia honestly. Bears have been wrong on the business for a long time. Bulls have been right on demand. But a stock can be attached to a great company and still become difficult if investors price every bottleneck as permanent.
The real risk is dependency, not demand
The quarter does not show weak demand. It shows the opposite. The risk is dependency.
The AI ecosystem is depending on Nvidia for compute. It is depending on Nvidia networking to keep clusters efficient. It is depending on CUDA, libraries, inference software, reference systems and the upgrade path from Blackwell to Vera Rubin. Customers may complain about cost, but the data suggests they are still buying because the alternative is falling behind.
That dependency is Nvidia's moat. It is also the market's risk map.
If every frontier lab, cloud provider, AI cloud, sovereign project and enterprise buyer needs Nvidia, the company gets extraordinary pricing power and visibility. If those customers eventually decide they must reduce that dependency, Nvidia's growth can stay large while the multiple changes. That distinction matters. Nvidia does not have to collapse for the stock to re-rate. It only has to stop looking like the only route into scalable AI production.
The competitors are not asleep. Custom silicon from hyperscalers will keep improving. AMD will keep pushing. Startups will keep attacking inference economics. Customers will keep negotiating. Governments will keep inserting policy into supply chains. Power availability and data center construction timelines will remain constraints.
None of that invalidates the quarter. It simply means the quarter's strength is also a measurement of how much pressure the rest of the ecosystem has to reduce single-vendor dependence.
The edge story is early, but not decorative
Edge Computing revenue was $6.4 billion, up 10% sequentially and 29% from a year earlier. That is small next to Data Center, but it is not irrelevant.
The edge bucket includes data processing devices for agentic and physical AI, including PCs, game consoles, workstations, AI-RAN base stations, robotics and automotive. The reason this matters is not that edge revenue will overtake Data Center soon. It matters because Nvidia's long-term pitch is that AI does not stay inside hyperscale data centers. It moves into factories, vehicles, robots, telecom infrastructure, workstations and local devices.
If that happens, Nvidia's platform logic gets stronger. The same company that powers training clusters can sell into inference, robotics simulation, industrial systems and physical AI deployment. That is a much broader surface than the old gaming-plus-data-center story.
Still, investors should be disciplined. Edge revenue is real, but Data Center is still the engine. The stock will move on the AI factory buildout, not on a distant robot dream unless the robot dream becomes a visible revenue line.
The verdict
This was a spectacular quarter. There is no need to pretend otherwise. Revenue growth, Data Center scale, networking growth, gross margin, free cash flow and Q2 guidance all point in the same direction: AI infrastructure demand is still expanding faster than skeptics expected.
But the better article is not the one that shouts "Nvidia wins again." The better article asks what Nvidia is becoming.
It is becoming the company that monetizes AI demand at the point where ambition hits physical constraints. Need more training? Nvidia. Need inference throughput? Nvidia. Need cluster networking? Nvidia. Need enterprise AI systems? Nvidia. Need a migration path from Blackwell to Vera Rubin? Nvidia. Need sovereign AI infrastructure? Nvidia wants to be there too.
That is why the quarter matters. Nvidia is not just selling chips into the AI boom. It is becoming the ledger where the AI boom shows up in dollars.
The risk is that the ledger is now so large that every future entry has to be extraordinary. The company proved the AI factory buildout is still accelerating. The stock now has to prove the market has not already capitalized too much of that future.
For now, the honest read is simple: Nvidia gave bulls the numbers they needed and gave skeptics a better version of the question they were already asking. This is not a demand problem. It is a duration problem. The company is winning. The debate is how long the win can keep compounding at this scale.
This article is research and commentary, not investment advice. Earnings reactions can reverse quickly, especially in high-expectation stocks like Nvidia.
Jazib Zaman founded TECHi in 2010 and is chief executive of its publisher, TechAbout LLC. He writes about AI infrastructure, semiconductors and the companies financing them, from Palantir's growth expectations to the toolmakers behind AI memory demand. He is a former member of the Forbes Technology Council.
Comments