Connect the dots and you get one story told twice. NVIDIA’s FY2021 10-K widens the language from last year: the GPU is now a platform “to create platforms for scientific computing, artificial intelligence, or AI, data science, autonomous vehicles, or AV, robotics.” The list grew, and so did the money: total revenue of $16.7 billion, up from $10.9 billion the year before.
That is a 53% increase, and the number is the proof the language is not just marketing. A year ago this site read NVIDIA’s 10-K as a forward bet on AI in the data center. The new filing is the first annual report where the revenue growth starts to back the rhetoric, even if the document still does not promise where the ceiling is.
Follow both the money and the technology. The expanded use-case list, AV and robotics now named alongside AI and data science, is the company telling the SEC it intends the same chip to address several compute-hungry markets at once. That is a deliberate breadth play: one architecture, many workloads, each of which independently needs the parallel math GPUs are good at.
For a general reader, the takeaway is that “AI chip company” undersells what the filing describes. NVIDIA is positioning the GPU as general infrastructure for any field that has become a data-and-compute problem, and 2021’s revenue jump is the first hard evidence that customers are buying the pitch.
As a contemporaneous marker, early 2021: the AI-platform thesis stops being a claim and starts being a growth rate. Whether 53% is a one-year surge or the start of a curve is exactly what the next filing will tell us, and we will read it the same way, against the document. Filing data and the evidence index via EdgarBeast.