NVIDIA's data-center business reaches $89 billion in its fiscal second quarter

NVIDIA reported $96.2 billion of revenue for the fiscal second quarter ended July 26, 2026. Revenue increased 18% from the preceding quarter and 106% from the same period a year earlier. The figures show that accelerated-computing demand translated into another large sequential increase as well as a year-over-year doubling.

This is a completed financial period, unlike capacity announcements that describe future deployment. The results provide a measured view of shipments and services recognized in the quarter, although they do not reveal unit volumes or average prices for individual GPU systems.

Data Center generated $89.0 billion, up 117% year over year. That segment accounted for most of total revenue and makes the quarter primarily an AI-infrastructure result rather than a broad consumer-chip story. NVIDIA also reported a 75.0% GAAP gross margin.

Gross margin reflects the company-wide mix and costs; it should not be interpreted as the margin on a particular accelerator. The release includes both GAAP and non-GAAP tables, but both gross-margin measures were 75.0% for the quarter. Comparisons should use the same accounting basis across periods.

For the third fiscal quarter, NVIDIA expects $108.0 billion of revenue, plus or minus 2%. The company explicitly assumes no China Data Center compute revenue in that outlook. This makes the geographic assumption part of the forecast rather than a hidden adjustment.

An outlook is management guidance, not recognized revenue. Supply, customer timing, regulation and product transitions can change the result. The exclusion also means any permitted China compute sales would sit outside the central assumption, while restrictions or demand changes elsewhere could still affect the total.

The scale and growth of Data Center indicate that spending on AI training and inference remains the dominant driver of NVIDIA's business. A 117% year-over-year increase is also a signal to cloud providers, networking vendors, memory suppliers and power developers whose capacity plans depend on accelerator deployment.

At the same time, rapid growth raises a difficult comparison base. Future percentage gains will be measured against increasingly large quarters. Investors and customers should distinguish demand indicators from supply commitments and watch whether infrastructure around the GPU—networking, memory, facilities and energy—keeps pace.

The next checkpoint is the $108.0 billion third-quarter revenue range and the stated absence of China Data Center compute revenue in that plan. Product milestones around Vera Rubin, networking and system deployments will matter because revenue depends on complete platforms reaching customers.

Useful follow-up measures include Data Center growth, gross margin, inventory, customer concentration and cash flow. A single quarter confirms current scale but cannot establish the duration of an infrastructure cycle. The company's next report will show whether sequential expansion continues near the rate embedded in guidance.

The release does not disclose revenue by individual GPU generation, cloud customer or country. It also does not state how much of the data-center result came from compute systems versus networking and related products.

The outlook range cannot resolve policy risk around China or predict customer deployment schedules. The supported conclusions are the reported quarter, segment scale, margin and management forecast. Claims about future market share, stock performance or guaranteed infrastructure demand would go beyond the evidence.