Nvidia's Q2 2027 earnings show a company operating at a scale that would have looked implausible during the early stages of the artificial-intelligence boom.
The chipmaker reported second-quarter fiscal 2027 revenue of $96.2 billion, up 106% from a year earlier and 18% from the previous quarter. Data Center revenue reached $89.0 billion, an increase of 117% year on year.
GAAP operating income was $63.7 billion, while net income reached $59.7 billion. Those numbers show an AI infrastructure cycle that is not merely continuing — it is becoming larger.
Data centres now dominate the company
Nvidia was once best known as a gaming-chip business. It is increasingly difficult to describe it that way: Data Center accounted for more than nine-tenths of quarterly revenue.
Hyperscalers, frontier AI laboratories, sovereign projects and enterprises are all competing for accelerated computing capacity.
Nvidia's advantage is not simply one chip. Its hardware, networking and software ecosystem has become deeply embedded in how developers build large-scale AI systems, which makes switching harder and helps sustain premium margins — a concentration also visible in index-level AI exposure.
Gross margin remains extraordinary
Nvidia reported a GAAP gross margin of 75.0%. For a hardware company at this scale, that level remains exceptional and demonstrates the pricing power created by scarcity and product differentiation.
Margins are also one of the figures investors will watch most closely from here. As competition expands and the company ships increasingly complex systems, production and deployment costs can rise.
Nvidia expects gross margin to remain around 74% in the third quarter. Even a small change matters when revenue approaches $100 billion per quarter.
The next generation is already in production
Chief executive Jensen Huang said the Vera Rubin platform is now in full production. That matters because the AI infrastructure market cannot depend indefinitely on one generation of accelerators.
Customers spending tens of billions of dollars need confidence that the performance curve will continue improving, and each platform transition creates execution risk.
Nvidia needs manufacturing partners, memory suppliers such as those driving the HBM memory cycle, networking equipment and data-centre capacity to scale alongside its own products.
China remains a notable absence
The company's outlook excludes China data-centre compute revenue, a reminder that geopolitical restrictions still shape the addressable market.
Nvidia can produce extraordinary growth while operating with limited access to one of the world's largest technology markets. If restrictions tighten, that remains a risk; if some access eventually returns, China represents potential upside.
Investors therefore need to distinguish operational demand from political availability — a theme running through semiconductor price-target revisions this year.
The valuation debate is now about duration
Nvidia has demonstrated that AI demand can generate exceptional revenue and profits. The harder question is how long growth at this scale can continue.
Global technology companies and AI laboratories are committing hundreds of billions of dollars to infrastructure, and those investments need eventually to generate returns for Nvidia's customers.
For now, the quarter provides little evidence of a slowdown. Investors are no longer asking whether the AI infrastructure boom is real; they are asking how long it can remain this strong.