Nvidia China Revenue Drops to Zero, Huang Confirms

Nvidia's China Business Hits Zero as Export Controls Bite Hard

Nvidia CEO Jensen Huang confirmed in May 2026 that the company's business in China has 'completely stalled,' marking a dramatic collapse from what was once a 95% stranglehold on the country's high-end AI chip market. Speaking to the bipartisan Special Competitive Studies Project (SCSP) — a Congressional initiative focused on technology competition — Huang stated bluntly: 'In China, we have now dropped to zero.'

The revelation, while not entirely unexpected, underscores the profound and accelerating impact of US export controls on the global AI chip landscape. It also raises critical questions about whether America's strategy of technological containment is achieving its goals — or simply reshaping the competitive battlefield in ways no one fully anticipated.

Key Takeaways

• Nvidia's China market share for high-end AI chips has fallen from 95% to 0% over roughly 2 years
• Huang made the disclosure during a Congressional interview with the SCSP in May 2026
• Despite the China losses, Nvidia projects $78 billion in quarterly revenue, a 77% year-over-year increase
• North American tech giants and Middle Eastern sovereign AI deals have offset Chinese demand
• The H20 chip ban was the final blow that eliminated Nvidia's remaining China-compliant product line
• Wall Street initially shrugged off the China losses when Huang first flagged them in October 2025

From 95% to Zero: The Timeline of Nvidia's China Collapse

Nvidia's dominance in China's AI chip market was once so complete that it bordered on monopoly. At its peak, the Santa Clara-based chipmaker controlled an estimated 95% of China's high-end AI accelerator market. Every major Chinese tech company — from Baidu to Alibaba to Tencent — relied on Nvidia GPUs to power their AI ambitions.

The erosion began in October 2022 when the Biden administration first introduced sweeping export restrictions targeting advanced semiconductors. Those initial controls forced Nvidia to create downgraded chips like the A800 and H800, specifically designed to comply with US performance thresholds while still serving Chinese customers.

But the rules kept tightening. Subsequent rounds of restrictions in 2023 and 2024 closed loopholes and lowered performance ceilings further. The final blow came with the ban on the H20 chip — Nvidia's last China-compliant AI accelerator — which effectively eliminated the company's ability to sell any meaningful AI hardware into the world's second-largest economy.

Wall Street Shrugs While Geopolitics Reshapes the Market

When Huang first disclosed in October 2025 that Nvidia's Chinese market share had cratered to zero, the financial markets barely flinched. The reason was simple: Nvidia's revenue pipeline was overflowing from other sources.

North American hyperscalers — Microsoft, Google, Amazon, and Meta — were engaged in what can only be described as a GPU arms race, collectively spending hundreds of billions on AI infrastructure. Meanwhile, Middle Eastern sovereign wealth funds had emerged as major buyers, with countries like Saudi Arabia and the UAE investing heavily in 'sovereign AI' capabilities.

Nvidia management projected quarterly revenue of $78 billion for the period, representing approximately 77% year-over-year growth. Those numbers made the China losses look almost irrelevant on a spreadsheet. But the strategic implications extend far beyond quarterly earnings.

The Strategic Paradox: Containment vs. Competition

The complete shutdown of Nvidia's China business creates a fascinating strategic paradox for US policymakers. On one hand, the export controls have achieved their stated objective: cutting off China's access to the world's most advanced AI training hardware.

On the other hand, the policy has created powerful incentives for China to develop domestic alternatives. Companies like Huawei, with its Ascend series of AI chips, and startups backed by massive state funding are racing to fill the vacuum Nvidia left behind.

• Huawei's Ascend 910C is reportedly being deployed across Chinese data centers as a direct Nvidia replacement
• SMIC, China's largest chipmaker, continues to advance its manufacturing capabilities despite equipment restrictions
• Chinese AI labs have become world leaders in training efficiency, learning to do more with less powerful hardware
• The domestic chip ecosystem has received over $100 billion in state subsidies and investment commitments
• Open-source AI models from Chinese labs like DeepSeek have demonstrated competitive performance with Western counterparts

The irony is not lost on industry observers: by forcing Chinese companies off Nvidia hardware, the US may have accelerated exactly the kind of technological self-sufficiency it sought to prevent.

What This Means for the Global AI Chip Market

Nvidia's China exit has ripple effects that extend well beyond the two countries directly involved. The global AI chip market is now effectively bifurcating into two separate ecosystems — one Western-led and one Chinese — with significant implications for the entire technology industry.

For Nvidia investors, the near-term impact appears manageable. The company has successfully redirected its supply to customers willing to pay premium prices, and demand from Western cloud providers shows no signs of slowing. The $78 billion quarterly revenue projection speaks for itself.

For competing chipmakers, the picture is more nuanced. AMD, Intel, and emerging AI chip startups like Cerebras and Groq face the same export restrictions, meaning none can capitalize on Nvidia's China exit. The opportunity instead flows to domestic Chinese players operating outside the reach of US regulations.

For global AI developers, the bifurcation means increasing fragmentation in hardware ecosystems, software frameworks, and potentially even AI model architectures. Applications built for Nvidia CUDA may not easily port to Huawei's ecosystem, and vice versa.

The Broader Implications for US-China Tech Competition

Huang's testimony before the SCSP carries weight beyond Nvidia's balance sheet. The Special Competitive Studies Project, co-chaired by former Google CEO Eric Schmidt, has been instrumental in shaping Washington's approach to technology competition with China.

The complete zeroing out of Nvidia's China business provides Congress with a clear data point: export controls work as a blunt instrument. Whether they work as a strategic tool for maintaining long-term technological advantage is a far more complex question.

Several factors complicate the picture:

• China's AI capabilities continue to advance despite hardware limitations, suggesting software innovation can partially compensate for chip restrictions
• The loss of Chinese revenue reduces Nvidia's R&D budget relative to what it could be, potentially slowing the pace of American chip innovation
• Allied nations like Japan, South Korea, and the Netherlands face pressure to align with US restrictions, creating diplomatic friction
• The emergence of alternative supply chains for AI hardware could eventually undermine the effectiveness of export controls entirely
• Chinese companies are increasingly turning to cloud-based workarounds and offshore computing arrangements to access restricted technology

Looking Ahead: What Comes Next

The trajectory of Nvidia's China business — from total dominance to total exclusion — may serve as a preview of broader technological decoupling between the world's two largest economies. As AI becomes increasingly central to economic competitiveness and national security, the pressure to restrict cross-border technology flows is unlikely to ease.

For Nvidia specifically, the path forward relies on continued explosive demand from Western and allied markets. The company's upcoming Blackwell Ultra and Rubin architectures are expected to command even higher prices, potentially offsetting volume losses from China through premium pricing.

However, Jensen Huang himself has acknowledged the risks. In previous statements, he warned that overly aggressive export controls could 'cause different parts of the world to develop their own technology' — a scenario that would permanently shrink Nvidia's addressable market.

The AI chip war between the US and China has entered a new phase. With Nvidia's China revenue at zero, the question is no longer whether decoupling will happen, but how far it will go — and who ultimately benefits from a world divided into competing technological blocs. For the global AI industry, the answer to that question will shape investment decisions, research priorities, and competitive dynamics for decades to come.

As Washington prepares its next round of technology policy decisions, Huang's stark testimony serves as both validation and warning: the export controls achieved their immediate objective, but the long-term consequences remain deeply uncertain.