Nvidia reported fiscal third-quarter revenue of $115.8 billion on November 17, up 103 percent from a year earlier and 20 percent from the prior quarter. Data center sales did most of the work, and the company guided to about $130 billion for the next quarter. That outlook assumes no data center compute revenue from China, a constraint that will shape the company's growth for some time.

The Numbers

Data center revenue reached a record $107.6 billion, up 110 percent year over year and 21 percent sequentially. That segment made up roughly 93 percent of total revenue, which underscores how concentrated the business has become around AI infrastructure. GAAP gross margin was 74.8 percent and non-GAAP gross margin was 74.9 percent.

Profitability also rose sharply. GAAP diluted earnings per share were $2.70, and net income was $65.3 billion, according to figures summarized from the release. The company returned $28.1 billion to shareholders during the quarter through buybacks and dividends and added $150 billion to its repurchase authorization.

Vera Rubin Ramps Up

The most important operational detail is the start of shipments for the Vera Rubin NVL72 rack system. Nvidia says it ramped production and began shipping the racks after what it describes as the fastest product ramp in its history. Systems are already running at partners including CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius, with another partner expected to bring systems online soon.

A fast ramp matters because it determines how quickly customers can convert orders into capacity. For cloud providers that have committed large budgets to AI buildouts, the timing of rack deliveries affects their own revenue and capital plans. In that sense, the product ramp is as important to the outlook as the headline revenue figure.

The Guidance

For the fourth quarter, which includes an extra week, Nvidia expects revenue of about $130 billion, plus or minus 3 percent. That outlook excludes data center compute revenue from China. The company also expects gross margin between 71 and 72 percent on both GAAP and non-GAAP bases, a step down from the 74.8 percent GAAP margin in the third quarter. Operating expenses are projected at roughly $11.0 billion on a GAAP basis.

Margin guidance is where investors will look for signs of pressure. Newer, faster-ramping hardware often carries different cost structures, and a lower margin guide is a reminder that scale does not automatically preserve profitability. The company's guidance does not single out one cause in the summary reviewed here, so the margin trajectory bears watching in future reports.

Shareholder returns were also large. The company returned $28.1 billion to shareholders during the quarter through repurchases and dividends and added $150 billion to its buyback authorization. Those returns sit alongside guidance for higher revenue, so the next quarter's gross margin and the pace of Vera Rubin shipments will be the signals to watch in future reports.

Why It Matters

Nvidia's numbers are a read on the AI infrastructure cycle. Revenue more than doubling in a year, with data center sales at 93 percent of the total, shows how much customer spending is flowing into accelerated computing. It also shows how dependent the company, and the wider supply chain, is on a small group of large buyers.

The China assumption is a second signal. By guiding without data center compute revenue from China, Nvidia is making clear that its outlook does not rely on sales there. Any future change in export rules would be upside to that guidance, but investors should not count on it until it happens.

For enterprise buyers, the message is that AI infrastructure supply is still tight and that the biggest vendors are prioritizing volume and ramp speed. Teams planning AI budgets should watch whether cloud pricing and availability keep pace with this build-out, because the cost of compute is ultimately set by how fast suppliers can deliver.

Sources