Nvidia has stopped all Hopper-series AI chip shipments to China as new U.S. export controls took effect November 19, wiping $2.4 billion in quarterly Chinese revenue from its forward guidance.
Nvidia Halts All H200 AI Chip Shipments to China as U.S. Export Rules Take Effect
Nvidia has stopped all Hopper-series AI chip shipments to China, effective today, as new U.S. Commerce Department export controls on advanced semiconductors went into force. The halt erases one of Nvidia's largest remaining revenue streams from the Chinese market — and the company's own guidance now assumes zero future China sales.
What Happened
Nvidia interim CEO Colette Kress disclosed the freeze during the company's Q3 earnings call. According to Technology.org, Kress confirmed that Nvidia has ceased all Hopper-series shipments to China and that forward revenue guidance carries no assumption of Chinese sales resuming.
The timing is not coincidental. New Commerce Department rules restricting the export of advanced AI accelerators to China took effect on November 19, 2026 — today. The H200 GPU, the flagship chip in the Hopper line, is Nvidia's most capable datacenter AI processor currently in wide production.
In the most recent quarter, Nvidia shipped $2.4 billion worth of Hopper processors to China — a significant slice of total datacenter revenue. That number now goes to zero.
Context
The U.S. government has steadily tightened semiconductor export controls on China since 2022, citing national security concerns over advanced AI hardware being used for military applications. Earlier restrictions blocked the most powerful chips — the A100 and H100 — which prompted Nvidia to develop China-specific, reduced-performance variants like the H800 and A800 to stay within export thresholds.
Get this in your inbox.
Daily AI intelligence. Free. No spam.
The new rules extend restrictions further down the performance spectrum, effectively closing the workaround. The H200, which delivers higher memory bandwidth than the H100, was one of the last high-tier chips Nvidia could still ship to Chinese buyers without special licensing.
Chinese cloud providers, AI labs, and enterprise customers had been stockpiling Hopper chips in anticipation of exactly this restriction. Demand spiked sharply in the months leading up to November 19, contributing to the $2.4 billion final-quarter figure.
The Numbers
- $2.4 billion — Hopper chip revenue from China in Q3 2026 (the last quarter before the freeze)
- $0 — China revenue assumed in Nvidia's forward guidance
- H200 — Nvidia's top Hopper-series chip, now blocked for Chinese export
- Nvidia's total Q3 datacenter revenue was not disclosed in this report; the China share represents a meaningful portion
What This Means
For Nvidia: The company absorbs an immediate multi-billion-dollar annual revenue hole. Nvidia has historically been exposed to China representing 20–25% of datacenter revenue in prior quarters. The question now is how fast demand from U.S., European, and Middle Eastern customers can offset the loss — and whether Blackwell-generation chips, Nvidia's next architecture, will face similar restrictions from the start.
For Chinese AI companies: Firms like Baidu, ByteDance, Alibaba, and Chinese datacenter operators now must accelerate domestic alternatives. Huawei's Ascend 910C processor is the leading domestic option, though it trails Nvidia's chips on raw performance benchmarks. Domestic production capacity is also constrained.
For the broader AI market: Expect demand pressure on every Nvidia chip that isn't restricted — U.S. and allied buyers will compete harder for available Blackwell supply. Pricing leverage shifts further toward Nvidia in unrestricted markets.
For chipmakers watching from the sidelines: AMD's MI300X and MI400 series face the same export control regime. AMD has not issued comparable guidance yet, but the same restrictions apply.
What to Watch
Whether China retaliates through trade restrictions on materials critical to semiconductor manufacturing — rare earth elements and specialty chemicals — where it holds significant supply leverage. Also watch Nvidia's next earnings call for a revised geographic revenue breakdown that reflects the post-freeze baseline.
Did this help you understand AI better?
Your feedback helps us write more useful content.
Get tomorrow's AI briefing
Join readers who start their day with NexChron. Free, daily, no spam.