US Demands 15% Cut of Nvidia and AMD AI Chip Sales
The Trump administration secured a reported 15% commission on Nvidia and AMD AI GPU sales to China as part of export license approvals for stripped-down chips like Nvidia’s H20 and AMD’s MI308. The deal, reportedly agreed with Nvidia CEO Jensen Huang, could bring roughly $2 billion a year to the U.S. and sharpens tensions over tech export policy and supply-chain risk.
What happened
The Trump administration has reportedly required Nvidia and AMD to pay a 15% commission on AI GPU sales to China as part of export approvals, according to The New York Times and The Financial Times. The move followed a period of tightened export controls and comes after both companies received licenses to ship modified, lower-capability chips such as Nvidia’s H20 and AMD’s MI308.
Reports say Nvidia CEO Jensen Huang struck the arrangement with President Trump days before the Commerce Department issued the licenses. The New York Times estimates the levy could net roughly $2 billion a year for the U.S. government.
Why this is unusual
Experts call the arrangement "highly unusual." It's not typical for a government to take a fixed revenue cut tied to private sales as a condition of export approval. The deal sits alongside a string of unconventional interventions from this administration — from proposed joint ventures for social apps to threats of sweeping tariffs aimed at reshaping chip manufacturing.
Key implications
- Chipmakers: immediate revenue and pricing pressure; contract renegotiations likely.
- Supply chains: incentives to reshore manufacturing or localize production to avoid future levies or tariffs.
- Customers in China: potential distrust after reports of hardware backdoors and pricing shifts; buyers may look to domestic alternatives.
- Policy signal: sets a precedent where export approvals can carry direct fiscal conditions, blurring trade policy and revenue generation.
What companies should consider now
- Run revenue-impact scenarios that include levies, tariffs, and changed demand in China.
- Audit products and supply chains for compliance and national-security scrutiny.
- Prepare communications for customers and regulators to manage trust and explain technical mitigations.
This is also a reminder that geopolitical risk can move faster than corporate planning cycles. A single policy shift can alter contracts, margins, and supplier choices almost overnight.
Bottom line
The reported 15% cut on AI chip sales to China reframes export licenses from a compliance checkpoint into a potential revenue lever for governments. For companies, the immediate task is pragmatic: quantify the impact, shore up compliance and supply-chain options, and communicate clearly with customers. For policymakers and market watchers, this move raises questions about how trade tools will be used in the accelerating competition over AI capability.
QuarkyByte’s approach is to combine scenario modeling, policy analysis, and supply-chain mapping to turn headline risks into actionable roadmaps. Companies and agencies navigating these changes need both technical audits and economic simulations to decide where to invest, what contracts to revise, and how to keep strategic projects on track.
Keep Reading
View AllApple Tests Siri That Can Operate Apps by Voice
Bloomberg: Apple tests a Siri that uses App Intents to perform actions across apps—search, edit, post, and send using voice commands.
US Secures 15% Cut on AI Chip Sales to China
Nvidia and AMD will pay 15% of revenue from high-end AI chip sales to China as part of U.S. license deals, raising trade and security questions.
Kentucky Town Leverages AI to Bridge Divides
Bowling Green used AI for a virtual town hall of 8,000 residents, synthesizing 4,000 ideas into 2,370 widely backed policy proposals that crossed political lines.
AI Tools Built for Agencies That Move Fast.
QuarkyByte can model the financial, regulatory, and supply-chain impact of this 15% levy for chipmakers, cloud providers, and government buyers. We run tailored scenarios to quantify revenue shifts, compliance exposure, and mitigation strategies so leaders can make data-driven decisions now.