China’s Nvidia “ban”: a speed bump or a roadblock?

The Financial Times reported China’s Cyberspace Administration (CAC) has instructed its tech giants, including Alibaba (NYSE: BABA), and TikTok’s parent company ByteDance, to stop buying Nvidia’s China-specific AI GPUs (RTX Pro 6000D). Despite the negative news, data suggest Nvidia’s long term growth remains largely intact without China. If sentiment knocks Nvidia’s share price lower, it could create a better entry point for investors who believe in the company’s multi-year growth story with the booming of artificial intelligence (AI).

What actually happened?

After the Biden administration tightened advanced computing export controls in 2022-23, Nvidia created a detuned H20 chip to comply. In April 2025, the Trump administration imposed new restrictions blocking H20 exports to China. Nvidia responded with the lower-powered RTX Pro 6000D. Recently, the CAC reportedly instructed Chinese tech giants to stop buying even this chip, sending Nvidia’s share price lower.

Why the ban doesn’t break the thesis

1. China is neither a huge slice nor in the base case.

According to Nvidia’s latest 10-K, contribution from China (including Hong Kong) only accounted for about 13% of FY2025 (US$17.1bn of US$130.5bn) and 6% of FY26Q2 (US$2.8bn of US$46.7bn) revenue. Following the tighter US export controls, CEO Jensen Huang told investors that “going forward, our forecasts will not include the China market”. In other words, the company already treats China as zero in its base case revenue and profit outlook; any revenue from China would be on the upside.

2. AI workloads won’t stop at borders.

Enterprise AI adoption keeps accelerating globally. McKinsey’s 2025 State of AI finds 71% of organisations now use generative AI (gen-AI) in at least one business function, up sharply from 33% adoption in 2023. Amongst larger firms, more than 52% of organisations with US$500m+ revenue have established a dedicated team to drive gen-AI adoption.

3. Hyperscale capacity is booming.

Analysts estimate Amazon AWS (NASDAQ: AMZN), Microsoft Azure (NASDAQ: MSFT) and Google Cloud (NASDAQ: GOOG) together captured roughly 60–65% of global cloud infrastructure spending in 2Q 2025, growing at 22% year-on-year, fuelled by AI workloads. Oracle (NYSE: ORCL) OCI is also scaling quickly from a smaller base. This hyperscale build out forms the computational backbone, driving strong demand for Nvidia’s chips.

4. Automotive: a growing AI play.

Although revenue from the automotive segment is still in its early stages, Nvidia’s DRIVE platform (Orin today, Thor next), its AI-based in-vehicle computing platform designed to enable smart driving, has already deeply embedded across global manufacturers, including Chinese EV makers Nio and Xpeng. More automakers, including GM and Toyota, are also buying Nvidia’s auto chips.

5. Agentic AI: the next growth wave.

“Agentic AI”— autonomous software agents identified by Gartner as a top enterprise trend for 2025–26. It is set to become another powerful demand flywheel, both for training AI models and for delivering practical use cases. By enabling AI systems to operate independently and handle complex tasks, agentic AI is positioned to drive the next wave of enterprise adoption and turn the concept of “AI agents” into real world productivity gains.

US still leads the AI stack

Nvidia may dominate the headlines from time to time, it’s not just Nvidia benefiting from the AI story. Broadcom (NASDAQ: AVGO) expects AI chip revenue to go from US$5.2bn in Q3 to US$6.2bn in Q4 this year, reflecting the robust hyperscale demand. AMD (NASDAQ: AMD) continues to grow its data centres and chip sales despite China export headwinds. On the infrastructure side, Microsoft (NASDAQ: MSFT), Amazon (NASDAQ: AMZN), Alphabet (NASDAQ: GOOGL) and Oracle (NYSE: ORCL), all US companies, collectively command the lion’s share of global AI computation power. US leadership across the AI stack is very robust.

Portfolio angle: a “pure-play US tech” exposure that includes the critical leaders

Australian investors have increasingly turned to ETFs to gain exposure to the US market, given the ASX 200’s relatively low weighting in technology stocks. For those seeking exposure to AI growth but unsure which companies to pick, ETFS Technology ETF (Cboe: WWW) offers a diversified way to participate. The ETF captures a broad range of leading US tech companies and is well-positioned to benefit from the AI dominance.

Classification that fits the market. WWW tracks the Solactive United States Technology Index which uses Factset’s RBICS classification system. This framework better captures companies that we all think of as “technology”, including Meta (NASDAQ: META) and Alphabet (NASDAQ: GOOG), but are excluded from many Information Technology sector ETFs because of the classification used by major indices. For example, Meta and Alphabet have been classified in the Communication Services sector in GICS since 2018.