Nvidia’s $96 Billion Quarter Reframes the AI Spending Debate

The quarter in one sentence

Nvidia’s latest results turned an already large company into a larger macro signal: revenue for the fiscal second quarter ended July 26, 2026 reached $96.2 billion, up 18% from the previous quarter and 106% from a year earlier, while the company projected $108.0 billion in revenue for the third quarter of fiscal 2027, plus or minus 2% [1]. That is not simply an earnings beat; it is a statement that demand for AI computing capacity remains ahead of supply and that the largest buyers of chips, systems and data-center infrastructure are still committing capital at extraordinary scale [2].

The core of the report was the data-center business. Nvidia said Data Center revenue reached $89.0 billion in the quarter, up 18% sequentially and 117% from a year earlier, meaning the company’s AI infrastructure engine generated well over nine tenths of total revenue [1]. Edge Computing, which includes chips and platforms tied to PCs, robotics, vehicles and other local AI uses, contributed $7.2 billion, up 13% from the previous quarter and 27% year over year [1]. The mix matters because it shows Nvidia is not being valued primarily as a traditional semiconductor supplier; investors are treating it as the operating system, toll collector and capacity allocator for a broad AI buildout.

A beat, but against extreme expectations

The headline numbers cleared Wall Street’s published expectations. AP reported that Nvidia’s $96.22 billion in revenue exceeded analysts’ average forecast of $92.27 billion, and Reuters reported that the third-quarter forecast of $108 billion was above the $104.19 billion average estimate compiled by LSEG [2][3]. Non-GAAP diluted earnings per share were $2.22, while GAAP diluted earnings per share were $2.46, and GAAP net income was $59.7 billion for the quarter [1].

Those results would be spectacular for almost any company, but Nvidia is now judged against an unusually high bar. Reuters described an initial after-hours reversal that turned into a gain of more than 4% as investors focused on the growth roadmap, while AP said the stock finished the regular session down 1.6% and was up 12.4% for the year before the post-close move [2][3]. In other words, the market reaction was not about whether Nvidia is growing; it was about whether the company can continue growing fast enough to justify the expectations attached to the AI trade.

The $108 billion guide is the new benchmark

The most consequential number may be the forecast, not the quarter just reported. Nvidia’s third-quarter outlook calls for $108.0 billion in revenue, plus or minus 2%, with GAAP and non-GAAP gross margins expected at 74.0%, plus or minus 50 basis points [1]. Reuters reported that analysts had expected an adjusted gross margin of 74.77%, making the margin guide a point of scrutiny even as the revenue forecast beat consensus [2].

The guidance also comes with an important geographic caveat. Nvidia said it is not assuming any Data Center compute revenue from China in the third-quarter outlook [1]. Reuters described China sales as uncertain and noted that the company’s outlook excludes data-center chip sales from China, while AP also highlighted that China compute revenue is not included in the forecast [2][3]. That means the $108 billion target is built without a China upside case, but it also means any continuing export-control friction remains a material swing factor.

Huang’s message: compute has become revenue

Chief executive Jensen Huang framed the results as a change in how the economy thinks about computing capacity. Nvidia quoted Huang as saying that AI is doing useful work, that tokens are productive and profitable, and that “compute is revenue” [1]. The phrase is important because it links Nvidia’s hardware sales to the business models of AI labs, cloud providers, software companies and enterprises trying to convert inference into paid services.

Huang also pointed to a broader base of demand, saying that the buildout is no longer driven by one lab but by multiple frontier labs, startups, open-model developers and physical AI use cases [1]. That argument addresses one of the biggest concerns around Nvidia: concentration. If growth depends on a narrow group of buyers, a pause in spending by a few cloud or AI customers could hit the company quickly. If demand is widening across model builders, sovereign infrastructure, enterprises and robotics, the cycle becomes deeper and harder to time.

Supply is still part of the story

The results suggest demand is not the only constraint. AP reported that CFO Colette Kress told analysts the company’s growth outlook would be closer to double, based on customers’ own forecasts, if Nvidia were not constrained by supply needed to meet chip production demand [3]. That point helps explain why the company’s forecast can be both enormous and still viewed by some investors as less than the most optimistic scenario.

The supply issue is also visible in the product roadmap. Nvidia said its Vera Rubin platform is ramping into full production, with racks running at partners including CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius [1]. The company also said Spectrum-6 switch systems tied to the Vera Rubin platform are arriving across gigascale AI factories, and it described new systems, software and security initiatives aimed at AI factories and agentic AI workloads [1]. The strategic message is that Nvidia is selling not only GPUs but the surrounding rack-scale architecture, networking, CPUs, software and operational playbooks needed to deploy AI capacity.

The financing debate is getting louder

The same report that showcased Nvidia’s operating strength also intensified debate over how the AI boom is being funded. Nvidia announced strategic partnerships to establish independent compute financing platforms with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR, with the goal of mobilizing more than $500 billion of third-party capital for AI infrastructure over time, subject to definitive agreements [1]. Reuters said the arrangement has increased scrutiny of Nvidia’s role in the AI boom’s financing ecosystem after the company agreed to guarantee some deals under a tie-up with six major financial institutions [2].

This is the central tension in the Nvidia story now. On one side, financing partnerships can accelerate the buildout of data centers, power access and compute leasing models, which can pull more demand into Nvidia’s ecosystem. On the other side, investors are alert to circularity: the risk that a supplier helps finance customers who then buy more from that same supplier. Axios reported that Nvidia’s earnings arrived after a series of investments and capital maneuvers designed to strengthen the AI economy and its customer base, including a recent deal to backstop an OpenAI data center [4].

Big Tech spending remains the demand floor

The demand backdrop still looks formidable. Reuters reported that Microsoft and Meta, two of Nvidia’s major customers, have reinforced expectations that Big Tech will spend more than $730 billion on AI infrastructure this year, compared with about $400 billion last year [2]. Reuters also reported that Nvidia and Amazon Web Services are expanding their partnership, with AWS set to deploy an additional 2 million Nvidia GPUs across Amazon’s global infrastructure in 2027 and 2028, according to Kress on the earnings call [2].

At the same time, Reuters noted that a growing share of technology-company spending is going toward in-house chip efforts aimed at reducing reliance on Nvidia’s costly and supply-constrained processors [2]. That is the competitive counterweight. Nvidia still has the clearest demand signal in AI infrastructure, but its largest customers are also among the companies most capable of trying to design around it.

What investors should watch next

Three indicators now matter most. First is the third-quarter revenue trajectory: hitting $108 billion would imply that Nvidia can keep compounding from a far larger base than skeptics expected [1]. AP reported that reaching the target for the August-to-October period would translate into roughly 89% revenue growth from a year earlier [3]. Second is gross margin, because memory costs, Rubin ramp costs and supply-chain complexity can determine how much of the revenue boom turns into earnings [2]. Third is the quality of demand, including whether revenue is increasingly supported by a diversified customer base rather than financing-supported hyperscale projects.

Nvidia has made the case that AI compute is becoming a new production input for the digital economy. The latest quarter supports that case with numbers: $96.2 billion in revenue, $59.7 billion in GAAP net income, $89.0 billion from Data Center, and a $108.0 billion guide for the current quarter [1]. But the same scale that makes Nvidia extraordinary also makes every caveat more important. China is excluded from the outlook, supply is tight, margins are being watched closely, customers are experimenting with their own chips, and the financing model behind AI infrastructure is under scrutiny [1][2][3].

The result is a paradox: Nvidia looks stronger than ever operationally, yet the questions around it are also larger than ever. The company has not merely beaten estimates; it has become a test of whether the AI infrastructure cycle can remain self-sustaining as it moves from billions to hundreds of billions of dollars in quarterly and annual commitments [2][4].

Sources from the last 72 hours

1. [1]NVIDIA Announces Financial Results for Second Quarter Fiscal 2027Aug 26, 2026, 8:20 PM UTC
2. [2]Nvidia forecasts quarterly revenue above estimates, shares riseAug 26, 2026, 8:28 PM UTC
3. [3]Nvidia projects 70% revenue growth in 2028Aug 26, 2026, 9:51 PM UTC
4. [4]Strong AI chip demand powers Nvidia's Q2 results past 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.