Nvidia's $96bn quarter shows AI spending has no ceiling yet

Nvidia's results confirm what its customers have been signalling for months: the AI build-out is still accelerating, and Nvidia is still the toll collector. The chipmaker reported revenue of $96.2 billion for the quarter ended 26 July 2026, up 18% on the previous quarter and 106% on a year earlier, comfortably beating the roughly $92.1 billion analysts had expected.

Data-centre revenue, the business that now defines the company, came in at $89.0 billion, up 117% year on year. Gross margin held at 75.0% on both GAAP and non-GAAP measures, and net income reached $59.7 billion. Non-GAAP earnings per share of $2.22 beat the $2.09 consensus, the fifth straight quarterly beat, according to Nvidia's results announcement.

The company returned $26.0 billion to shareholders through buybacks and dividends, leaving about $99.0 billion under its existing buyback authorisation. It will pay a $0.25-per-share dividend on 1 October. Debt, though, has jumped to $33.4 billion from $8.5 billion in January, a sign Nvidia is funding its own expansion as aggressively as it is supplying everyone else's.

Nvidia's guidance points to $108.0 billion in revenue for the current quarter, plus or minus 2%, and it is assuming zero data-centre compute revenue from China. That figure lands just days after customers were warned server prices would rise more than 15% on memory costs, and as Nvidia commits to backing up to $105 billion of financing for OpenAI's Ohio data centre.

The China exclusion in the guidance is the number to watch: Nvidia is telling the market it can hit $108 billion without that market at all, which either understates the upside if access returns or flags how large a structural gap Beijing's restrictions have carved out. Either way, the company is no longer hedging its forecasts on a resolution.

The financing commitment to OpenAI and the rising debt load point to a shift in how Nvidia extracts value from the boom: it is becoming a lender and infrastructure backer to its own biggest customers, not just their supplier. That deepens the dependency on both sides, and leaves Nvidia's balance sheet more exposed to a slowdown in AI capital spending than it has been in years.