Nvidia Halts China Chip Exports as U.S. Restrictions Tighten

Nvidia’s efforts to export advanced AI chips to China have hit a standstill, following aggressive U.S. government restrictions that now require special licenses—none of which have been granted. The effective ban has major implications for Nvidia’s operations in China and its broader AI strategy.

The chipmaker disclosed in a recent SEC filing that it expects to take a hit of up to $5.5 billion in its first-quarter earnings. This includes losses tied to the halted export of its H20 chips, unsold inventory, and previous production costs.

The H20 Chip at the Center of the Ban

The H20 was Nvidia’s answer to earlier export rules—designed to stay within technical thresholds allowed under U.S. trade restrictions. However, in 2024, the U.S. imposed even stricter rules aimed at curbing China’s access to high-performance AI accelerators. Now, even these downscaled chips require a U.S. export license.

Despite technically allowing exports with permission in the form of a license, the U.S. has not approved any licenses, creating a de facto ban.

Strategic and Financial Fallout

• Financial Impact: Nvidia warned investors that it could write down billions in inventory that can no longer be sold to Chinese clients.

• Market Share Concerns: Analysts warn the ban may backfire. With the H20’s performance already lagging behind domestic Chinese alternatives, the export halt could drive customers toward companies like Huawei, accelerating China’s self-reliance in AI hardware.

• Shifting Investment: In response, Nvidia has committed $500 billion over the next four years to build out U.S.-based AI infrastructure—including supercomputers and data centers—signaling a pivot toward domestic expansion.

Huawei Steps In as U.S. Blocks Nvidia, AMD Shipments

With Nvidia’s AI chip exports to China effectively frozen, Huawei Technologies is rapidly filling the gap. According to Reuters, the company has begun limited shipments of its new 910C chip, with mass production set to begin next month.

The 910C, built by combining two 910B processors, reportedly matches the performance of Nvidia’s H100 chip, a 2022 model that remains widely used in AI applications. While not as advanced as the now-restricted H20, Huawei’s 910C is gaining traction as a viable domestic alternative.

China’s Semiconductor Manufacturing International Corp. (SMIC) is said to be producing these chips, though analysts warn that SMIC may face challenges scaling production to meet demand. Still, Huawei’s upcoming Ascend 920 chip, expected later this year, is rumored to rival the H20 in performance.

“If SMIC can surmount these issues and U.S. restrictions remain in place, eventually new offerings from Huawei could permanently constrict Nvidia's opportunity in China,” wrote Wedbush Securities analyst Matt Bryson.

Meanwhile, AMD has also been affected by the U.S. trade ban. The company expects to take a charge of up to $800 million due to unsellable inventory tied to its MI308 chip.

On the stock market, both firms took a hit Monday:

• Nvidia shares fell 4.5%, closing at $96.91

• AMD stock dropped 2.2%, ending the day at $85.56

With Nvidia down nearly 28% year to date, investors are growing increasingly wary of long-term exposure to the Chinese AI market.

What This Means

The latest developments mark a turning point in the escalating tech rift between the U.S. and China. What began as targeted export controls has evolved into a high-stakes reshaping of the global AI chip landscape. For Nvidia and AMD, the financial toll is immediate—but the strategic cost could be longer-lasting.

China is not waiting on the sidelines. Huawei’s emergence as a viable competitor—paired with aggressive domestic chip development by SMIC—signals a clear push toward semiconductor self-sufficiency. If these efforts prove successful, U.S. companies may find themselves edged out of one of the world’s largest AI markets, not just temporarily but permanently.

At the same time, the Trump administration’s tech policy is forcing U.S. chipmakers to refocus on domestic innovation and infrastructure. Nvidia’s $500 billion investment plan reflects both a retreat and a pivot—one that could strengthen U.S. AI leadership at home while ceding ground abroad.

This isn’t just a commercial dispute—it’s a defining moment in the geopolitical struggle for AI dominance. This moment may mark the beginning of a long-term decoupling in the global AI hardware race—where speed, self-reliance, and political alignment will determine who shapes the next frontier.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.