When Nvidia CEO Jensen Huang initially told the Financial Times that China would “win the AI race” before softening his stance, it crystallised a predicament that’s been years in the making. The world’s most valuable chipmaker now finds itself caught between two superpowers, each wielding the Nvidia AI chip ban as a weapon in a broader technological cold war—and the company’s attempt to please both sides may ultimately satisfy neither.
From dominance to zero: A market collapse
The numbers tell a stark story. Speaking at a Citadel Securities event in October, Huang revealed that Nvidia’s share of China’s AI accelerator market has collapsed from roughly 95% to zero, with the company now assuming no revenue from China in its forecasts. This isn’t just a revenue hiccup—China previously represented between 20% and 25% of Nvidia’s data centre revenue, a segment that generated more than US$41 billion in its most recent financial results.
The latest blow came this week when sources claimed that the White House informed federal agencies it will not permit Nvidia to sell its latest scaled-down AI chips to China, specifically the B30A chip designed to train large language models. Despite Nvidia providing samples to Chinese customers and reportedly working to modify the design, the Trump administration has drawn a hard line.
But Washington’s restrictions represent only half of Nvidia’s problem. Beijing has issued guidance requiring new data centre projects receiving state funds to use only domestically-made AI chips, with projects less than 30% complete ordered to remove all installed foreign chips or cancel purchase plans.
It’s a pincer movement that leaves Nvidia with virtually no room to manoeuvre.
The lobbying game: Too much, too late?
Huang has long argued that maintaining China’s dependence on American hardware serves US interests. His logic? Keep Chinese AI developers hooked on Nvidia’s ecosystem, and America retains technological leverage.
Following meetings with President Trump in July, it appeared Huang’s lobbying had worked, with Washington agreeing to ease some chip curbs under a plan where Nvidia and AMD would pay the US government 15% of their Chinese revenues.
That optimism proved short-lived. Beijing has since shut Nvidia out of the market through a national security review of its chips, with Huang stating the firm’s market share has been reduced to zero. The irony is palpable: while Huang lobbied Washington to allow more sales to China, Beijing was simultaneously building barriers to keep Nvidia out.
When Huang contrasted China’s pro-industry energy subsidies with what he described as excessive Western regulation, it revealed the fundamental tension in Nvidia’s position. The company needs a favourable policy from both capitals, but operates in an environment where pleasing one increasingly means antagonising the other.
The cost of technological nationalism
This isn’t merely a corporate problem—it’s reshaping the global AI landscape. China’s ban would eliminate foreign chipmakers like Nvidia from a significant portion of the market, even if a deal is agreed to allow the resumption ofadvanced chip sales to China.
Meanwhile, Chinese companies have over US$100 billion in state funding for AI data centre projects since 2021, creating a massive captive market for domestic alternatives.
The policy whiplash has real consequences. Following Trump’s meetings with Chinese President Xi Jinping, highly anticipated trade talks yielded no concessions from either side on chip policy, with top US officials rallying against Trump’s initial consideration of Huang’s request to allow sales of new AI chips to China.
An Nvidia spokesperson’s response to the latest restrictions was telling Reuters: “zero share in China’s highly competitive market for datacenter compute, and do not include it in our guidance”. It’s a public acknowledgement of defeat wrapped in corporate speak.
China’s calculated response
Beijing’s moves reveal a strategy that extends beyond mere retaliation. China has discouraged local tech giants from purchasing advanced Nvidia chips over security concerns this year, while showing off a new data centre powered solely by domestic AI chips. The message is clear: foreign dependence is a vulnerability to be eliminated, not managed.
The Chinese government is carving out market share for domestic chipmakers ranging from Huawei Technologies to smaller players like Shanghai-listed Cambricon and startups including MetaX, Moore Threads, and Enflame.
While these companies struggle to match Nvidia’s performance and software ecosystem, they’re getting exactly what they need most: time, money, and a protected market to mature.
The impossible balance
Nvidia’s predicament exposes a broader truth about technology in an era of great power competition: the middle ground is disappearing. Companies can optimise for American national security priorities or Chinese market access, but increasingly not both.
Huang expressed concerns that the West was being held back by “cynicism” and excessive regulation, contrasting this with China’s energy subsidies aimed at lowering costs for local developers using domestic chips. But this comparison misses the point.
The question isn’t whether China’s industrial policy is more effective—it’s whether Nvidia can operate in an environment where technology has become inseparable from geopolitics. The B30A saga illustrates the futility of technical compromises.
Even a chip deliberately neutered to comply with US export controls finds no approval from Washington, while Beijing increasingly views any foreign chip as a strategic vulnerability. Nvidia could design a thousand variants, each weaker than the last, and still find itself shut out by one capital or the other.
What comes next?
In the short term, Nvidia faces a stark reality: the company now assumes 0% revenue from China in all forecasts, with Huang stating, “If anything happens in China… it will be a bonus”. This conservative guidance protects the stock but signals that management sees no near-term resolution.
The real question is whether this represents a temporary freeze or a permanent fracture. While the move helps boost sales of domestically developed chips, it also risks widening the US-China gap in AI computing power, as US tech giants continue spending hundreds of billions on data centres powered by Nvidia’s most advanced chips.
For Nvidia, the path forward likely involves doubling down on markets where geopolitics align with business—the US, Europe, and friendly Asian nations. The China dream, at least in its previous form, appears over. Huang’s softening of his “China will win” comments reflects this new reality. America might not win by keeping China dependent on its chips, but Nvidia certainly loses by being caught in the middle.
The Nvidia AI chip ban—from both directions—represents more than export controls or industrial policy. It’s evidence that in the AI race, there won’t be neutral suppliers. Technology companies will increasingly be forced to choose sides, and those who hesitate will find the choice made for them.
Nvidia’s plunge from 95% to zero market share in China took just months. The question now is whether Washington and Beijing will leave any space for global tech companies to operate at all.
(Photo by OpenAI and Nvidia plan $100B chip deal for AI future)
See also:
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The post Nvidia AI chip ban: Can tech giants navigate a geopolitical zero-sum game? appeared first on AI News.
source https://www.artificialintelligence-news.com/news/nvidia-ai-chip-ban-china-market-share/
