NVIDIA Q2 FY2027 Revenue Hits $96.2B, Up 106% YoY; $108B Outlook Assumes Zero China Data Center Compute Revenue

NVIDIA reported Q2 FY2027 results on August 26, 2026: revenue of $96.221 billion, up 106% year over year, with $89.0 billion from Data Center. The Q3 outlook of $108.0 billion assumes no Data Center compute revenue from China at all.

NVIDIA earnings data-center AI-infrastructure GPU Vera-Rubin

NVIDIA reported results for the second quarter of fiscal 2027, ended July 26, 2026, on August 26¹. Revenue came in at $96.221 billion, up 18% sequentially and 106% year over year¹. Data Center accounted for $89.0 billion of that, a 117% increase from a year earlier¹.

For gauging how far real demand for AI infrastructure extends, the quarter’s figures and the guidance that follows them are useful material — and the assumption baked into that guidance says as much as the number itself.

The quarter’s figures

Gross margin was 75.0% on both a GAAP and non-GAAP basis, and diluted earnings per share were $2.46 GAAP and $2.22 non-GAAP¹. GAAP net income was $59.688 billion, up 126% year over year; on a non-GAAP basis it was $53.954 billion, up 118%¹. GAAP operating income reached $63.734 billion, up 124% from a year ago¹.

Sequential comparisons diverge depending on the measure. Revenue rose 18% and operating income 19%, but GAAP net income rose only 2% from the prior quarter¹. Non-GAAP net income over the same period rose 18%¹.

On capital returns, NVIDIA sent approximately $26.0 billion back to shareholders through buybacks and cash dividends during the quarter, with roughly $99.0 billion remaining under its repurchase authorization at quarter end¹. The next quarterly dividend of $0.25 per share is payable October 1, 2026, to shareholders of record as of September 10, 2026¹.

An outlook built without China

Third-quarter revenue is expected to be $108.0 billion, plus or minus 2%¹. The sentence attached to it deserves more attention than the figure: NVIDIA states it is not assuming any Data Center compute revenue from China in that outlook¹.

The $108.0 billion level, in other words, is assembled without any Data Center compute revenue from China in the arithmetic. Gross margins are expected at 74.0% GAAP and non-GAAP, plus or minus 50 basis points, with operating expenses of approximately $9.2 billion GAAP and $9.0 billion non-GAAP¹. The full-year fiscal 2027 tax rate is expected between 16.0% and 18.0%, excluding discrete items and material changes to the company’s tax environment¹.

Jensen Huang, NVIDIA’s founder and CEO, said in the release that AI has reached its inflection point and that compute is now revenue, adding that where a single lab drove the buildout this time last year, multiple frontier labs are now scaling in parallel alongside an open-model ecosystem and physical AI coming online¹. That is the company’s chief executive speaking, not an independent market assessment.

Announcements from the summer, gathered in one list

The Data Center highlights section of the release collects items announced individually over recent months. The Vera Rubin platform is ramping into full production, with racks running at partners including CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure, and Nebius¹. Vera, which the company describes as the first CPU built for AI agents, was announced on August 24 as a chip SpaceXAI will deploy. The inference accelerator Groq 3 LPX entering full production appears in the same list¹.

Land and power show up as line items too. Securing land, power, and shell capacity at the PORTS-Pike Technology Campus in Ohio through a partnership with SB Energy was announced on August 17¹. On the site development side, a minority investment in Cloverleaf Infrastructure came to light on August 21.

On financing, one item describes strategic partnerships with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to establish independent compute financing platforms mobilizing over $500 billion of third-party capital for AI infrastructure over time¹. It carries the note subject to definitive agreements, so it is not a concluded arrangement¹.

Two numbers from the same day, supply side and buyer side

Also on August 26, AWS and NVIDIA announced an expansion of their collaboration. AWS plans to deploy an additional 2 million Blackwell Ultra, Rubin, and Rubin Ultra GPUs across its global infrastructure in 2027-2028, AI factories included². At NVIDIA GTC 2026, AWS had announced plans to add more than 1 million GPUs starting in 2026; demand has exceeded those expectations since, according to the release².

The results and that deployment plan are separate announcements, but for a reader they are two faces of one question: how much easier will it get to obtain the compute that runs a given workload? That next-quarter guidance is stacked up without any Data Center compute revenue from China, and that a cloud provider is putting numbers on additions two years out, both feed into that estimate.

Remarks from the earnings call and the market’s reaction are not part of the press releases reviewed here, and this article stays within what those releases state.