Nvidia Posts $81.6B Q1 FY2027 Revenue, Guides $91B With Zero China Sales

Photo by Mariia Shalabaieva on Unsplash

Nvidia reported $81.6 billion in revenue for the April quarter, raised its dividend 25-fold, and authorized another $80 billion in share repurchases, all while telling investors the next quarter will be guided as if China Data Center revenue does not exist. The Nvidia Q1 FY2027 earnings results landed with a $2.8 billion beat against Street consensus, and a forward guide that still managed to top expectations by more than $4 billion despite zeroing out one of the company's historically largest geographies.

The paradox driving this print is simple. Demand from US hyperscalers is absorbing every Blackwell board Nvidia can ship, fast enough to mask the structural loss of an entire country from the model.

The Numbers Behind Nvidia Q1 FY2027 Earnings Results

Revenue hit a record $81.6 billion, up 20% sequentially and 85% year-over-year, according to Nvidia's official Q1 FY2027 press release. Data Center alone contributed $75.2 billion, a 92% year-over-year jump and the largest single segment the company has ever reported in a quarter.

Profitability scaled with the top line. Kiplinger's live coverage put net income at $58.3 billion against an analyst estimate of $42.9 billion, a $15.4 billion beat that arrived even as the company funded a steep ramp in operating expenses.

Gross margin is the quieter headline. IndMoney's earnings recap noted that Nvidia held gross margin at 75% through Blackwell's full production ramp, defying analyst worries that the transition would compress margins by several points. A flat margin through a node and architecture shift is the kind of result that quietly reprices a stock, even when the headline beat is the part that gets quoted.

Why China Data Center Revenue Is Now Zero in the Guide

The most consequential sentence in Nvidia's filings this quarter is not in the income statement. It is in the guidance footnote. The 8-K summary on StockTitan confirms that the $91.0 billion Q2 FY27 revenue guide, plus or minus 2%, with a 75.0% non-GAAP gross margin, explicitly excludes any Data Center compute revenue from China.

For context, China Data Center revenue was $4.6 billion in the comparable quarter a year earlier. That line item is now modeled at zero.

This is not a forecasting choice. It is the cumulative effect of US export controls that took the H20 off the table for Chinese customers, followed by Beijing's own pushback on the chip's national-security profile. Computer Weekly reported that the H20 was already restricted from China sales, and on the earnings call, Nvidia's CFO confirmed that no H200 shipments to China have occurred or are anticipated under current rules, per the transcript published by The Motley Fool. Until policy changes, that is a structurally permanent headwind, not a quarterly noise item.

The overlooked implication is this. Nvidia has effectively decoupled its forward model from any China upside, which means every dollar of Chinese demand that does eventually materialize, through licensing or a new export-compliant SKU, drops into a guide that has already been validated by the market without it. Investors are no longer paying for China revenue, which makes any return of it pure optionality.

The $80 Billion Buyback and the 25× Dividend

The Nvidia $80 billion stock buyback authorization, approved by the board effective May 18, 2026, carries no expiration date. It also stacks on top of an existing program. On the earnings call, the CFO clarified that the $80 billion is on top of $39 billion already remaining under the prior authorization, bringing total repurchase capacity past $119 billion.

The dividend move is sharper. Nvidia raised the quarterly cash dividend from $0.01 to $0.25 per share, a 25-fold increase. Per Kiplinger, the new $0.25 payout is payable June 26, 2026, to shareholders of record on June 4, 2026.

The Nvidia dividend increase 2026 still leaves the yield trivial against the share price, but the symbolism is what matters. Nvidia is signaling that free cash flow has crossed the threshold where capital return becomes a permanent line item, not an afterthought. The CFO told analysts the company plans to return roughly 50% of free cash flow to shareholders in FY2027, according to the call transcript. About $20.0 billion was already returned in Q1 alone through buybacks and dividends combined, per the 8-K.

A company guiding to 50% FCF return while also funding the largest capex-adjacent supply commitments in semiconductor history is a rare combination. It only works if the demand curve is as steep as Nvidia thinks it is.

The $119 Billion Supply Commitment Hiding in the Footnotes

Inventory rose to $25.8 billion and supply-related purchase commitments hit $119.0 billion, per the StockTitan filing summary. That second number is the one worth staring at.

$119 billion in supply commitments is a bet that the AI buildout does not slow in the next 18 months. If hyperscaler order patterns soften even modestly, Nvidia is left holding fab allocations, HBM contracts, and CoWoS capacity that it cannot easily unwind. IndMoney estimates combined hyperscaler AI capex for 2026 at roughly $725 billion, which is the demand wall the supply commitment is leaning against.

GAAP operating expenses also rose 52% year-over-year to $7.6 billion in Q1, with Q2 guided to about $8.5 billion GAAP and $8.3 billion non-GAAP. On the call, the CFO described full-year FY2027 opex growth as landing in the "upper 40% range," driven by R&D and internal AI productivity tooling, per the Motley Fool transcript. That is not a maintenance budget. It is a company spending like it expects the Rubin architecture ramp to require headcount and design work at a tempo most chipmakers cannot match.

A quieter detail from the call. The full-year tax rate was revised down to 16% to 18%, from 17% to 19%, reflecting a shift in the geographic mix of revenue. Translation: less China, more US and other allied markets, with a friendlier effective rate as a side effect.

The $1 Trillion Forecast and What It Implies for Competitors

Nvidia projects $1 trillion in combined Blackwell and Rubin architecture revenue from 2025 through calendar 2027, according to commentary on the earnings call. That is the company's own forecast, and it functions as a public commitment to investors more than a precise model.

Jensen Huang framed the demand environment in characteristically large terms. In the official press release, Huang called the AI factory buildout "the largest infrastructure expansion in human history." He also told analysts that "Agentic AI has arrived, doing productive work, generating real value and scaling rapidly across companies and industries," per Kiplinger.

For competitors, the $1 trillion number is the harder problem. A challenger like Cerebras can point to architectural advantages on specific workloads, as covered in how Cerebras' wafer-scale engine works, but selling against a vendor with a trillion-dollar three-year forecast already pre-sold to the largest cloud buyers is a different battle than competing on flops per watt. AMD's MI series and the custom silicon roadmaps at Google, Amazon, and Microsoft will need to demonstrate that they can take share at scale, not just at the edges.

The broader picture for AI Infrastructure is that Nvidia has now embedded its supply chain so deeply with TSMC, SK hynix, and the major hyperscalers that the question is no longer whether anyone catches the company on chips. The question is whether the demand cycle holds long enough for Rubin to ship into a market that still wants it.

Why NVDA Q1 2027 Earnings Did Not Move the Stock

NVDA Q1 2027 earnings produced a record print, a $4 billion guide beat, the largest buyback authorization in company history, and a 25-fold dividend increase. The stock fell roughly 1.5% in after-hours trading and was already down about 0.9% in the regular session, according to IndMoney.

The explanation came from outside the company. Capital.com senior analyst Kyle Rodda told Kiplinger that the result was "a garden variety beat" that was "well telegraphed following the very strong results from AI-hyperscalers earlier in the earnings season." When Microsoft, Meta, Alphabet, and Amazon all confirm record AI capex weeks before Nvidia reports, the beat is mathematically priced in by the time the press release crosses.

Analyst 12-month price targets now cluster around $272 to $276 per share per the same IndMoney recap. The market is not arguing with the numbers. It is arguing with how much more upside is left after the largest hyperscaler capex year on record has already been telegraphed.

What to Watch Next

The specific date that matters is August 27, 2026, the expected Q2 FY27 print. The Nvidia $91 billion revenue guidance is plus or minus 2%, which gives a band of roughly $89.2 billion to $92.8 billion. A beat inside that band is consensus. A meaningful beat above the band, with China still modeled at zero, would force analysts to raise full-year estimates again. A miss at the low end, particularly if accompanied by any softening in hyperscaler commentary, would test the $119 billion supply commitment thesis for the first time.

The second event to watch is any movement on US export policy toward China. Any licensed return of compute-class GPUs to Chinese data centers would land as pure upside against a guide that has already removed the geography entirely.

For a wider view of how the chip and capex picture is shifting around these numbers, the News archive and ongoing AI Industry coverage on AnIntent track the supply-chain and policy moves that will determine whether Rubin ships into the same demand wall Blackwell has enjoyed.