Nvidia’s $96.2B Quarter Turns AI Infrastructure Into a Revenue Machine

A quarter that resets the scale of AI demand

Nvidia’s second-quarter fiscal 2027 report is less a normal earnings release than a new benchmark for the economics of artificial intelligence infrastructure. The company reported revenue of $96.221 billion for the quarter ended July 26, 2026, up 18% from the prior quarter and 106% from the same period a year earlier [1]. GAAP net income was $59.688 billion, up 126% year over year, and diluted GAAP earnings per share reached $2.46, while non-GAAP diluted EPS was $2.22 [1].

The center of gravity is unmistakable. Data Center revenue was $89.023 billion, or roughly 92.5% of total sales, up 18% sequentially and 117% year over year [2]. Nvidia’s CFO commentary attributed the performance to the ramp of Blackwell Ultra infrastructure, with hyperscale revenue more than doubling from a year earlier and the AI Clouds, Industrial and Enterprise category rising 138% year over year [2]. In practical terms, the quarter confirms that the company’s revenue base is now overwhelmingly tied to the construction and operation of AI data centers.

Nvidia also gave investors a striking forward signal: it expects fiscal third-quarter revenue of $108.0 billion, plus or minus 2%, while assuming no Data Center compute revenue from China in that outlook [1]. If achieved, that would mark Nvidia’s first quarter above $100 billion in sales, a threshold that only recently would have seemed implausible for a semiconductor company.

The data center business is now the business

The most important number in the report is not simply the $96.2 billion top line, but the $89.0 billion generated by Data Center. Nvidia’s reworked market-platform disclosure shows Data Center split between $48.710 billion of hyperscale revenue and $40.313 billion from AI Clouds, Industrial and Enterprise, or ACIE [2]. That split matters because it shows demand is not confined to a handful of U.S. cloud giants; management says AI natives, enterprises, sovereign customers and hyperscalers using AI clouds are also driving the ACIE category [2].

AP reported that Nvidia’s Data Center segment includes AI data centers and factories as well as demand from hyperscalers such as Amazon, Meta and Google [3]. The same report said the top five hyperscalers’ capital spending is expected to reach nearly $800 billion this year and $1.3 trillion in 2027, according to comments from CFO Colette Kress [3]. That is the context in which Nvidia’s latest revenue growth should be read: its chips, networking, systems and software are being purchased as the foundation for a multi-year capital cycle.

Management’s language also sharpened the strategic framing. Jensen Huang said in the earnings release that AI had reached an “inflection point,” that tokens were becoming productive and profitable, and that “compute is revenue” [1]. The line is promotional, but it captures the model Nvidia is selling: customers are not buying GPUs as isolated components; they are buying the ability to generate, serve and monetize AI outputs at industrial scale.

Profitability remains extreme, but costs are rising

The profitability profile remains exceptional. Nvidia reported both GAAP and non-GAAP gross margins of 75.0% for the quarter [1]. Operating income reached $63.734 billion on a GAAP basis, more than doubling from the year-earlier quarter [1]. Free cash flow was $21.341 billion, according to the company’s reconciliation [1].

Yet the report also contained signs of the cost and balance-sheet demands created by this growth phase. GAAP operating expenses rose 55% year over year to $8.408 billion, while non-GAAP operating expenses rose 54% to $8.232 billion [1]. The CFO commentary said higher compute infrastructure and compensation and benefits costs drove the increase [2].

Nvidia also returned approximately $26.0 billion to shareholders through repurchases and dividends during the quarter and ended the period with about $99.0 billion remaining under its share repurchase authorization [1]. That capital return sits alongside rapidly expanding strategic commitments. The CFO commentary said supplier commitments increased from $119 billion last quarter to $279 billion, primarily tied to memory procurement [2]. Across supply and capacity, cloud service agreements, data center leases, equity investments and capital expenditures, future commitments totaled $366 billion as of July 26, 2026 [2].

The new risk: financing the ecosystem that buys the chips

The earnings story is no longer only about demand; it is also about the structure of that demand. Nvidia’s CFO commentary described arrangements intended to help select customers secure land, power and data center capacity, saying some AI clouds and model makers are growing faster than their balance sheets and long-term credit profiles can support [2]. That statement is central to the debate now forming around Nvidia: the company is not only selling into the AI buildout, it is helping organize pieces of the infrastructure and financing needed to make the buildout happen.

The most eye-catching disclosure concerns guarantees. Nvidia said it entered into August 2026 guarantees supporting land, power and shell buildout for approximately 4.25 gigawatts at SB Energy’s PORTS-Pike Technology Campus in Ohio, which is expected to host Nvidia infrastructure under 20-year leases to OpenAI, subject to limited exceptions [2]. The company said those guarantee obligations are capped at $105 billion and become effective in phases as conditions are met, with the first data centers expected in fiscal 2029 [2]. It also disclosed an option to provide phased credit support for about 3.8 additional gigawatts as the site scales [2].

Axios reported that Kress acknowledged criticism that such support could be viewed as circular financing, while arguing that Nvidia sees the arrangements differently because the companies involved have strong technology leadership and fast-growing usage [4]. This is now a core investor question: whether Nvidia is merely enabling customers through a temporary infrastructure bottleneck, or whether parts of the AI demand curve increasingly depend on Nvidia-backed capital structures.

Markets wanted more than a beat

The numbers beat expectations. AP reported that revenue of $96.22 billion surpassed analysts’ average forecast of $92.27 billion and that adjusted EPS of $2.22 exceeded the $2.09 consensus [3]. Axios said the company also exceeded S&P Capital IQ expectations of $92.1 billion in revenue and $51.2 billion in earnings for the period ended July 26 [4].

Still, market reaction was not simply euphoric. AP reported that Nvidia shares ended the regular session down 1.6% but rose 4.1% in after-hours trading following the earnings call [3]. Axios reported a similar post-call move, saying shares rose 4.4% after Kress delivered a projection for about 70% revenue growth in Nvidia’s next fiscal year [4]. That sequence shows the level of investor expectation around Nvidia: a historic quarter can initially be treated as insufficient unless management supplies an even stronger forward narrative.

The third-quarter guide helps that narrative. Nvidia’s forecast of $108.0 billion in revenue, plus or minus 2%, implies continued sequential growth even without assumed China Data Center compute revenue [1]. Constellation Research noted that such guidance would put Nvidia at a $432 billion annual revenue run rate [5]. That is not the same as a full-year forecast, but it illustrates how quickly the company’s quarterly scale is moving.

China is still a visible absence

China remains a constraint rather than a growth pillar in the latest outlook. The CFO commentary said shipments of Data Center Hopper products to China were less than 1% of Data Center revenue in the quarter [2]. Nvidia also said its Q3 outlook assumes no Data Center compute revenue from China [1]. For a company growing this fast, that absence is notable: the current trajectory is being driven largely without meaningful China data center compute contribution.

That cuts both ways. On one hand, it shows the depth of demand elsewhere, particularly from hyperscalers, AI clouds, enterprises and sovereign infrastructure programs. On the other hand, it underlines the role of export controls and geopolitics in Nvidia’s future mix. The company can clear $96 billion in quarterly revenue without relying on China data center compute, but any shift in licensing, restrictions or customer access could still affect investor assumptions.

What the quarter really says

Nvidia’s Q2 fiscal 2027 result confirms three things. First, AI infrastructure spending has not slowed in the company’s core markets; it has accelerated into a scale where quarterly sales near $100 billion are already in the rearview mirror. Second, the Data Center platform has become the economic engine of Nvidia, with Edge Computing revenue of $7.198 billion now playing a secondary role despite growing 27% year over year [2]. Third, the next phase of the story will be judged not only by chip demand, but also by supply commitments, customer financing, power access, lease exposure and the durability of AI model monetization.

For now, the operating momentum is extraordinary. Nvidia is pairing 106% annual revenue growth with 75% gross margins and an outlook for another step up in the next quarter [1]. But the company’s own disclosures show why the valuation debate is becoming more complex: the AI factory boom requires not just silicon, but land, power, memory, credit support and long-term commitments measured in hundreds of billions of dollars [2].

The quarter therefore strengthens both sides of the Nvidia debate. Bulls can point to record revenue, record Data Center sales, strong margins, a $108 billion Q3 guide and management’s claim that demand remains constrained by supply. Skeptics can point to soaring commitments, guarantees linked to future data center capacity, rising expenses and questions about whether Nvidia’s customer ecosystem is becoming financially intertwined. What is no longer in doubt is the scale: Nvidia has become the clearest financial proxy for the global AI infrastructure buildout, and Q2 fiscal 2027 pushed that proxy to a new level.

Sources from the last 72 hours

1. [1]NVIDIA Announces Financial Results for Second Quarter Fiscal 2027Aug 26, 2026, 8:00 PM UTC
2. [2]Nvidia projects 70% revenue growth in 2028Aug 26, 2026, 9:51 PM UTC
3. [3]Nvidia projects $432 billion annual revenue run rate on Q2 surgeAug 26, 2026, 12:00 AM UTC
4. [4]CFO Commentary on Second Quarter Fiscal 2027 ResultsAug 26, 2026, 8:00 PM UTC
5. [5]Strong AI chip demand fuels Nvidia’s Q2 results well beyond Wall Street’s expectationsAug 26, 2026, 8:40 PM UTC

AI-generated article based on recent web research, then preserved as a dated editorial snapshot.