On June 12, 2025, Nvidia CEO Jensen Huang confirmed in a CNN interview what analysts had suspected for months: the company’s revenue and profit forecasts no longer include China. Reuters and the South China Morning Post carried the report that day, and TechCrunch followed on June 13 under the headline “Nvidia to exclude China from its revenue and profit forecasts.”
This was not a retreat from China so much as a re-anchoring of Nvidia’s financial planning. With US export controls showing no endpoint, any forecast that still counted Chinese revenue would mislead investors and management alike. Huang chose to strip that variable out of the model entirely, and said so on the record rather than through a guidance footnote.
Forecasts formally exclude China
In the interview, Huang explained that Nvidia’s financial forecasts would henceforth assume zero contribution from the Chinese market. The logic is blunt: rather than carrying an unpredictable policy variable inside the model, plan on nothing and treat any recovery as upside. If Washington reverses course and Nvidia regains access to Chinese customers, he said, that would be a “bonus” rather than a planning assumption. In accounting terms, China-related revenue moved from “core assumption” to “upside scenario” in a single interview. That change in wording amounts to a new financial discipline: the company will no longer be judged each quarter on a market whose access depends on export licenses it does not control, and whose purchasing can be switched off by a single administrative notice.
The CNN interview: no bet on a policy reversal
The interview aired on June 12, 2025. Huang said he was not confident the US government would change its stance on chip export restrictions soon, TechCrunch reported. The significance of that admission is easy to miss: Nvidia stopped assuming that licenses would be granted, and stopped assuming that the rules would be relaxed. Reuters framed the story exactly that way — Nvidia would stop including China in its forecasts amid US chip export controls — and the South China Morning Post kept the same focus in its own headline. In effect, both wires treated the exclusion as a settled policy of the company rather than a passing complaint. The contrast with Huang’s public optimism about global AI demand is stark. Confidence in overall demand and confidence in China access now sit on two separate tracks, and only one of them feeds the official numbers.
H20 controls and the US$8 billion gap
The backdrop came in April 2025, when the US government required Nvidia’s H20 chips to obtain an export license before shipment to China. Nvidia subsequently said the licensing requirement would cost it roughly US$8 billion in second-quarter revenue — the April-to-June 2025 quarter — a figure TechCrunch cited in its June 13 report as the financial basis for Huang’s comments. Eight billion dollars is not a rounding error, but Huang’s handling of it is the instructive part. Rather than letting the market reprice Nvidia every time a license decision surfaced, he booked the loss as known, set the China assumption to zero, and defined everything above that as upside. That removes a recurring source of quarterly uncertainty and lets management concentrate on demand elsewhere. In practice, the guidance turned a geopolitical standoff into a plain line item. For anyone tracking Nvidia’s numbers, the China column is no longer a variable; it reads zero until further notice.
What it means for developers and product teams
For teams building AI products, the lesson from this episode is less about geopolitics and more about planning method: move uncontrollable policy variables out of the baseline scenario and treat reversal as upside, not as a dependency. Nvidia applied that discipline to its own financial model, publicly and at scale. Developers and product owners face the same structure of risk wherever export controls, tariffs, or data-sovereignty rules can rewrite market access with one administrative notice. A conservative baseline with clearly flagged upside is a better input for engineering roadmaps, procurement decisions, and hiring plans than an optimistic forecast that quietly depends on a license being granted. By June 2025, the largest AI chipmaker on the market had decided that distinction was worth stating on the record, in an interview, with an eight-billion-dollar number attached.
Sources
- TechCrunch: Nvidia to exclude China from its revenue and profit forecasts
- Reuters: Nvidia to stop including China in forecasts amid US chip export controls, CNN reports
- SCMP: Nvidia to stop including China in profit, revenue forecasts amid US chip export controls
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.