Nvidia Reports $130.5B FY2025 Revenue; Blackwell Ramp Begins
Summary
Nvidia reported fiscal year 2025 revenue of $130.5 billion, a 114% increase year-over-year, driven by explosive demand for AI infrastructure. Q4 revenue reached $39.3 billion, with data center revenue of $35.6 billion. The Blackwell GPU architecture generated $11 billion in Q4 alone, which Jensen Huang described as the fastest product ramp in company history.
What Happened
Nvidia's FY2025 results confirmed the company's dominance of the AI compute buildout. Full-year revenue of $130.5 billion more than doubled from $60.9 billion in FY2024, itself already a record. Data center remained the primary growth driver, contributing the vast majority of revenue as hyperscalers and AI labs competed intensely for GPU allocations.
The Blackwell architecture — Nvidia's next-generation GPU platform, successor to Hopper — generated $11 billion in Q4 revenue despite being in the early ramp phase. Nvidia had faced manufacturing delays on Blackwell during 2024, and the $11 billion Q4 contribution signaled that supply chains had stabilized. CEO Jensen Huang called it "the fastest product ramp in company history," a claim supported by the absolute revenue figures: $11 billion in a single quarter for a platform that had just begun shipping at scale.
Q4 data center revenue of $35.6 billion represented roughly 90% of total Q4 revenue of $39.3 billion, underlining how completely Nvidia's growth story had become an AI infrastructure story. Gaming, professional visualization, and automotive revenues were secondary contributors.
Gross margins remained above 70%, a level typically associated with software companies rather than semiconductor manufacturers, reflecting Nvidia's pricing power in a constrained supply environment where demand consistently outpaced availability.
Why It Matters
Nvidia's FY2025 results function as the financial foundation of the entire AI investment cycle. Every dollar invested in AI labs, every large language model trained, every inference server deployed ultimately flows through GPU purchases — and Nvidia holds roughly 80% market share in the high-performance AI compute segment. The $130.5 billion revenue figure is not just a company milestone; it is a proxy measure for the scale of the AI buildout.
The Blackwell ramp is particularly significant because it sets the supply ceiling for the next generation of AI model training. The labs and hyperscalers competing to train frontier models are constrained by Blackwell availability. That constraint shapes the competitive dynamics of the entire industry: who can train what, how fast, at what cost.
The financial profile — 70%+ gross margins on hardware, doubling revenue year-over-year — represents a supplier economics story with few historical parallels. Nvidia has captured the extraordinary margins usually associated with intellectual property licensing while selling physical chips. The durability of those margins depends on whether AMD, Intel, or custom silicon from Google, Amazon, and Microsoft can provide credible alternatives. FY2025 results suggest that competition remained insufficient to compress margins.