Key Highlights

  • Revenue from customers headquartered outside the US fell from 42% of total revenue in Q1 FY2026 to just 22% in Q1 FY2027, as export controls effectively removed China's data centre market from NVIDIA's addressable base.
  • China revenue (including Hong Kong) collapsed from $9.66 billion in Q1 FY2026 to $4.55 billion in Q1 FY2027, while US revenue surged from $25.7 billion to $63.8 billion in the same period.
  • No Data Centre Hopper product shipments to China occurred in Q1 FY2027, compared with $4.6 billion in the prior year quarter, the direct consequence of US export licensing restrictions.
  • Despite losing its largest non-US data centre market, NVIDIA grew total revenue by 85% and Data Centre revenue by 92% year-on-year, driven by US hyperscaler and enterprise Demand.
  • NVIDIA's own risk disclosures state that the company is effectively foreclosed from competing in China's data centre computing market, with the competitive gap benefiting domestic Chinese semiconductor companies.

 

When historians look back at the 2020s, they will likely identify two defining structural forces that reshaped the global technology industry: the emergence of artificial intelligence as a general-purpose technology, and the systematic decoupling of the US and Chinese technology ecosystems. NVIDIA's Q1 FY2027 filing, for the quarter ended April 26, 2026, contains within its geographic revenue disclosures a precise financial measurement of both forces operating simultaneously.

These are not abstract geopolitical observations. They are written in the numbers on page twenty of the 10-Q filing, and they tell a story of extraordinary consequence.

The Numbers That Tell the Story

In the first quarter of fiscal year 2026, NVIDIA's customers headquartered outside the United States accounted for 42% of total revenue. Taiwan contributed $7.6 billion. China, including Hong Kong, contributed $9.66 billion. Together, these two geographies represented a significant and growing share of the most important semiconductor company on the planet.

Twelve months later, the picture had fundamentally changed. Total revenue grew from $44.1 billion to $81.6 billion, an 85% increase. But the geographic composition of that growth tells a story that the headline figure obscures. US revenue grew from $25.7 billion to $63.8 billion — a 148% increase. China revenue fell from $9.66 billion to $4.55 billion — a 53% decline. Taiwan revenue grew modestly from $7.6 billion to $12 billion, primarily reflecting the Taiwan-headquartered cloud and hyperscale customers who purchase NVIDIA products for non-China deployments.

The share of revenue from customers headquartered outside the United States collapsed from 42% to 22%. NVIDIA did not lose the AI infrastructure race. It was administratively removed from competing in one of the world's largest AI Investment markets, and it still grew 85%.

The Architecture of Exclusion

Understanding why this happened requires tracing the specific regulatory timeline disclosed in NVIDIA's risk factors. The sequence began in August 2022, when the US government imposed export restrictions targeting China's semiconductor and supercomputing industries, specifically impacting the A100 and H100 chips. Then in October 2023, updated licensing requirements covered a broader range of products. In April 2025, the H20 — NVIDIA's China-specific degraded product designed to comply with existing controls — was itself subjected to a new licensing requirement. The result was a $4.5 billion charge in Q1 FY2026 for H20 excess inventory and purchase obligations.

By Q1 FY2027, the filing states plainly: 'we were effectively foreclosed from competing in China's data centre computing/compute market.' This is not hedged language from a legal department. It is a company telling its investors, in a document filed with the SEC, that the world's second-largest economy is no longer a viable market for its most important products.

What the Foreclosure Means for Global AI Development

The broader implication, which the NVIDIA filing identifies explicitly, is that the competitive exclusion from China has not been cost-free for NVIDIA's global position. The filing states that the company's effective foreclosure from the China market helped its competitors build larger developer and customer ecosystems to challenge NVIDIA worldwide. This is an acknowledgement that NVIDIA's CUDA software ecosystem, which is the deepest moat the company possesses, requires developer adoption to sustain its advantage. Every quarter that Chinese AI companies, researchers, and developers spend building on domestic or alternative hardware architectures is a quarter in which NVIDIA's ecosystem advantage erodes at the Margin.

Domestic Chinese semiconductor companies — Huawei's Ascend chips being the most visible example — have accelerated their development precisely because Chinese technology companies have had no alternative. The US export controls intended to deny China AI capability may, over a sufficiently long time horizon, have the unintended consequence of creating a credible second ecosystem for AI compute that did not previously exist.

The US Concentration Risk

For investors reading the filing, the geographic shift creates a concentration risk that deserves attention. With 78% of revenue now derived from US-headquartered customers, NVIDIA's financial performance is increasingly correlated with the Capital Expenditure decisions of a small number of US hyperscalers. The filing notes that in Q1 FY2027, three direct customers represented 21%, 17%, and 16% of total revenue. These three customers alone accounted for 54% of an $81.6 billion quarter.

The compounding machine that NVIDIA has built is, at this moment in its history, heavily dependent on the continued willingness and ability of a handful of US-based companies to spend at unprecedented rates on AI infrastructure. The depth of that commitment is visible in the $119 billion in Manufacturing commitments on NVIDIA's Balance Sheet. The question of how long that rate of spend continues, and whether it broadens to international customers as geopolitical tensions evolve, is the central variable in every forward projection.

Disclaimer: This article is for informational and educational purposes only. It does not constitute financial, investment, or any other form of advice. All data is sourced from NVIDIA Corporation's Form 10-Q for the quarter ended April 26, 2026, filed with the US Securities and Exchange Commission.