NVIDIA’s Data Center division reported $107.6 billion in revenue in the third quarter of fiscal 2027, accounting for almost 93% of the company’s $115.8 billion in quarterly sales. The figures published on November 17, 2026, underline how strongly NVIDIA’s business now depends on spending on artificial intelligence infrastructure.

What powered the segment’s 110% year-on-year expansion, how much of the improvement came from its latest Vera Rubin systems, and what do these results reveal about the AI investment cycle? The company’s financial disclosures and infrastructure announcements offer a clearer picture than the headline revenue number alone.

How large was NVIDIA’s Data Center business in Q3 FY2027?

NVIDIA reported Data Center revenue of $107.6 billion for the quarter ending October 25, 2026. This represented an increase of 110% from the same quarter a year earlier and 21% from the preceding quarter.

Using the company’s rounded figures, the division made up approximately 92.9% of total quarterly revenue. NVIDIA’s remaining $8.2 billion was reported in Edge Computing, its other named business segment in the earnings release.

The concentration is significant: changes in AI infrastructure budgets, hardware availability and the pace of new data-centre construction now have an outsized effect on NVIDIA’s consolidated results.

What drove the 110% increase in Data Center revenue?

NVIDIA identified the deployment of its Vera Rubin computing platform and the expansion of demand for accelerated computing as major elements of the quarter.

These systems are used in large-scale clusters designed for AI model training and inference, alongside networking and supporting infrastructure. As model deployment expands from experimentation into production workloads, customers need capacity to serve models as well as to train them.

Revenue growth, however, is a measure of sales recognised during the quarter. Announced capacity, prospective customer projects and future purchase plans are separate indicators; they should not all be treated as booked revenue.

What did NVIDIA confirm about Vera Rubin NVL72?

NVIDIA said it ramped production and began shipping Vera Rubin NVL72 during the quarter, calling it the fastest product ramp in the company’s history.

The earnings announcement identified racks operating at partners including CoreWeave, Google Cloud, Microsoft Azure, Oracle Cloud Infrastructure and Nebius. It also described wider adoption of Vera CPUs and the Rubin platform among AI infrastructure providers.

The release separately discussed planned deployments and future availability. These illustrate the scale of customer interest but do not identify the exact dollar amount of Data Center sales generated by each customer or each product.

NVIDIA did not provide a complete customer-by-customer or platform-by-platform breakdown of the $107.6 billion segment figure in its headline release. Precise revenue attribution to an individual cloud company would therefore require additional disclosure.

Why are cloud providers and AI factories so important?

Building an AI system at scale involves far more than buying individual accelerators. Operators must coordinate compute, high-speed networking, power, cooling and software across entire facilities.

NVIDIA’s strategy increasingly addresses this integrated deployment model, sometimes described as an AI factory. Its disclosures highlighted collaboration with major cloud platforms, specialised AI infrastructure providers and supply-chain partners.

The company also announced reference architectures, networking developments and plans to support new facilities. Such projects help explain the industry’s direction, although their eventual contribution to revenue depends on installation schedules, customer spending and execution.

How did Data Center compare with NVIDIA’s overall results?

For Q3 FY2027, NVIDIA reported:

Reported measureQ3 FY2027
Total company revenue$115.8 billion
Data Center revenue$107.6 billion
Edge Computing revenue$8.2 billion
Data Center growth, year on year110%
Data Center growth, quarter on quarter21%
Company GAAP gross margin74.8%

The 74.8% gross margin is a consolidated company figure, not a separately disclosed Data Center segment margin. Readers should distinguish overall profitability from the segment’s sales growth.

For the broader earnings announcement, including net income and the interim CEO arrangements, see our NVIDIA Q3 FY2027 earnings overview.

What does the $130 billion Q4 guidance suggest?

NVIDIA projected approximately $130 billion in total company revenue for Q4 FY2027, allowing for a 3% variation. It also said the fourth quarter includes an extra week and that its outlook assumes no Data Center compute revenue from China.

That forecast indicates management expects further overall expansion. It is not a specific forecast of $130 billion in Data Center revenue, nor a guarantee that demand or production will develop exactly as anticipated.

Margins are another factor. NVIDIA projected GAAP and non-GAAP gross margins of 71% to 72% for the next quarter, below the reported third-quarter level. Revenue growth and profitability must therefore be assessed together.

What should investors and technology buyers watch?

The next major signals will be the rate of Vera Rubin deployments, actual cloud capacity entering service, spending by the largest AI infrastructure providers and the relationship between sales growth and gross margins.

Industry-wide investment announcements can be substantial, but the timing of hardware delivery and revenue recognition determines when those investments appear in NVIDIA’s accounts.

The main finding is clear: $107.6 billion of NVIDIA’s Q3 FY2027 revenue came from Data Center, a segment growing more than twice as fast as its year-earlier base. The rollout of Vera Rubin and sustained cloud infrastructure demand are central to that performance. The question for subsequent quarters is how much of this pace can be sustained as installations scale and competition, power constraints and customer budgets evolve.