Nvidia & AMD to Pay 15% of China Chip Revenue to U.S. for Export Licenses

Meta Description: Nvidia and AMD have struck an unprecedented deal with the U.S. government—paying 15% of their China chip revenues to resume AI chip exports amid shifting tech-trade dynamics.

Summary: In a groundbreaking revenue-sharing agreement, Nvidia and AMD agreed to give 15% of their AI chip sales revenue in China to the U.S. government in exchange for export licences. This deal, brokered with the Trump administration, reignites China access but raises legal and policy concerns.

Introduction

Amid escalating tech tensions between the U.S. and China, semiconductor giants Nvidia and AMD have struck a rare deal—offering to hand over 15% of their China-origin chip sales as a condition for securing export licences. Approved by the Trump administration, the arrangement allows renewed sales of advanced AI chips like Nvidia’s H20 and AMD’s MI308. It’s a bold move that signals both desperation and innovation in navigating geopolitics.

Problem or Context

Export controls have long been wielded as tools of national security. Earlier in 2025, the U.S. halted AI chip exports to China, citing fears the technology could fuel military or surveillance advances. These curbs affected billion-dollar revenue streams for Nvidia and AMD and pressured both firms to juggle compliance, competitiveness, and investor expectations.

Meanwhile, China relaxed rare-earth export restrictions, and tensions over technology access intensified. The unique revenue-sharing agreement—with companies paying a cut of their proceeds to regain market entry—is unprecedented, raising constitutional, strategic, and ethical alarms.

Core Concepts Explained

This agreement rests on two core constructs:

• Export Licensing as Leverage: Traditionally, governments control tech exports via regulation. Paying a portion of revenue to lift restrictions blurs the line between licensing and taxation.
• Tech Truce via Trade Reciprocity: The U.S. appears to be leveraging trade access for both immediate revenue and maintenance of export control integrity—but critics argue it compromises the trust foundation.

Real-World Examples

In the SaaS world, firms often pay compliance or licensing fees, like for cloud infrastructure or enterprise authentication. Still, few commercial policies resemble paying a revenue cut to unlock market access.

In cybersecurity, specialized encryption hardware often sits under export restrictions. Yet, these restrictions are enforced strictly, without direct financial incentives from vendors.

In AI and blockchain, tokenized licensing—like paying a fee for registrar rights in a token, or pledging future royalties for IP use—has some parallels. However, this chip export scenario is far more expansive in its implications.

Use Cases and Applications

• AI-Driven Cloud Providers: Startups relying on Chinese Nvidia or AMD hardware can now regain access, fueling AI-as-a-Service growth.
• Embedded Systems Makers: Companies building IoT and industrial AI can tap Chinese supply chains again without breaching restrictions.
• Cross-Border R&D Collaboration: Research labs reliant on AI accelerators can resume partnerships and shared compute infrastructure.

Pros and Cons

Pros:

• Market Access Resumed: Nvidia and AMD regain billions in possible China revenue.
• Short-Term Revenue for U.S.: At a 15% cut on estimated $20B+ combined revenue, Washington may secure $3B+ in funds.

Cons:

• Possible Constitutional Violation: Some legal experts argue this duress-style levy behaves like an export tax—which the Constitution limits.
• Undermines Export Control Credibility: If revenue payments become loopholes for licensing, the U.S. risks eroding the effectiveness of trade restrictions.

Conclusion

This unprecedented deal between Nvidia, AMD, and the U.S. government speaks volumes about the geopolitical weight of semiconductors and AI capabilities. Analysts remain split: some welcome renewed Chinese market access, while others warn that turning tech restrictions into revenue schemes weakens national security frameworks.

What do you think? Does this deal offer pragmatic trade wrangling, or is it a slippery slope in tech policy? Share your thoughts below and let’s unpack the future of geopolitics in the AI era.