NVIDIA Corporation

NVIDIA Corporation is an American accelerated computing and AI infrastructure company that designs and sells graphics processing units (GPUs), data-center-scale computing platforms, networking equipment, and AI software, generating $215.9 billion in revenue in fiscal 2026. The company's CUDA development platform is used by over 7.5 million developers worldwide, and its data center products power AI workloads at every major cloud service provider, frontier model builder, and sovereign AI installation.

This is a story about an incumbent semiconductor franchise that became the central infrastructure supplier to the most consequential technology transition since the internet. NVIDIA's revenue more than tripled from $60.9 billion to $215.9 billion in two fiscal years, and the first quarter of fiscal 2027 delivered another 85% year-over-year acceleration — $81.6 billion in a single quarter, more than the company's entire annual revenue in fiscal 2024. The file turns on a single question: whether NVIDIA's platform dominance, which depends on an interlocking system of hardware leadership, software lock-in through CUDA, and an annual product cadence no competitor has matched, can endure the transition from hyperscaler-led infrastructure buildout to a broader, more contested AI market where customers, competitors, and governments all have strong incentives to reduce their dependence on a single supplier.

The scale is difficult to overstate. NVIDIA's $120.1 billion of net income in fiscal 2026 exceeds the annual revenue of all but roughly 50 public companies in the world. Its market capitalization of approximately $5 trillion implies that investors are pricing decades of AI infrastructure growth. The analysis that follows does not attempt to call the stock; it tries to make sense of the business beneath it — how the money is made, what could break the flywheel, and what signposts would tell you which way this goes.