NVIDIA posts $75.2B data-center revenue

NVIDIA reported first-quarter fiscal 2027 revenue of $81.6 billion on May 20, up 85% from a year earlier, with its Data Center business rising to a record $75.2 billion. The company said the quarter ended April 26, 2026. Chief Executive Jensen Huang said demand was being driven by the buildout of what he called “AI factories” and by the spread of agentic AI across industries. The figures are larger than the preliminary card summary suggested. NVIDIA’s official earnings release said total quarterly revenue was $81.6 billion, not $45.55 billion, while Data Center revenue was $75.2 billion, up 92% from a year earlier. The company also said it authorized an additional $80 billion in share repurchases and raised its quarterly cash dividend to $0.25 per share from $0.01. (investor.nvidia.com) ### Why does the Data Center number matter more than the headline revenue? Data Center accounted for most of NVIDIA’s quarterly sales. With $75.2 billion in that segment against $81.6 billion in total revenue, the business remains centered on the chips, networking and systems sold into AI training and inference workloads. NVIDIA has been framing that business less as standalone chip sales and more as infrastructure. (investor.nvidia.com) On its investor site, the company describes its evolution as moving “from chips to an AI infrastructure company,” and recent product announcements tied the next generation of systems directly to “AI factories” and agentic inference. ### What exactly did Jensen Huang say about “agentic AI” and “AI factories”? (investor.nvidia.com) Jensen Huang said in the earnings release that “the buildout of AI factories” is accelerating and that “agentic AI has arrived.” He said these systems are “doing productive work, generating real value and scaling rapidly across companies and industries.” (investor.nvidia.com) Those remarks match NVIDIA’s broader messaging this year. In March, the company said its Vera Rubin platform was designed for “every phase of AI, from pretraining, post-training and test-time scaling to agentic inference,” linking future hardware demand to workloads that run beyond one-off chatbot queries. ### Was this a step up from the prior quarter? (investor.nvidia.com) NVIDIA’s previous quarterly report, released on Feb. 25, showed total revenue of $68.1 billion and Data Center revenue of $62.3 billion for the fourth quarter of fiscal 2026. The latest quarter therefore added $13.5 billion in total revenue sequentially and $12.9 billion in Data Center revenue from the prior quarter. (investor.nvidia.com) The year-over-year growth rates also stayed high. NVIDIA said first-quarter revenue rose 85% from a year earlier, while Data Center revenue rose 92%. ### What does the report show about where AI spending is going? NVIDIA’s materials point to continuing spending on compute, networking and full-rack systems rather than a slowdown after the first wave of generative AI. (investor.nvidia.com) The company’s investor presentation highlights a projected $3 trillion to $4 trillion AI infrastructure spend by 2030, while its recent Dynamo software release said inference software is being optimized to raise throughput and lower token costs on Blackwell GPUs. (investor.nvidia.com) That framing comes from NVIDIA itself, not outside analysts. The company is telling investors that training demand is being joined by inference demand, especially for agent-based systems that require persistent, large-scale compute. ### What should readers watch next? April 29 was the date NVIDIA set for announcing its first-quarter call schedule, and May 20 was the date it released the results and webcast materials. (investor.nvidia.com) Investors can review the earnings release, presentation and replay through the company’s first-quarter fiscal 2027 event page. NVIDIA’s next scheduled corporate date on its investor calendar is the annual meeting of stockholders on June 24, 2026. (investor.nvidia.com 1) (investor.nvidia.com 2)