In a novel deal, the U.S. government granted NVIDIA export licenses for its H200 chips—advanced, but not cutting-edge—to markets like China. In return, NVIDIA pays a 25% fee on those sales. This establishes a new model where the government takes a direct revenue share from strategic tech companies in exchange for controlled market access.
Related Insights
China's pause on Nvidia H200 chip orders is not a permanent ban but a strategic move. The government aims to balance its immediate need for advanced AI chips with its long-term goal of fostering a competitive homegrown chip industry, preventing over-reliance on Western technology.
The decision to allow NVIDIA to sell powerful AI chips to China has a counterintuitive goal. The administration believes that by supplying China, it can "take the air out" of the country's own efforts to build a self-sufficient AI chip ecosystem, thereby hindering domestic firms like Huawei.
The US has reversed its strict chip controls on China. Instead of a complete ban, it now allows NVIDIA to sell advanced H200 chips but with a 25% tax, effectively turning a geopolitical restriction into a significant revenue stream for the US Treasury, estimated at $5 billion annually.
Allowing H200 chip sales gives China significant AI compute capability. This short-term revenue boost for NVIDIA won't alter China's long-term policy of reducing reliance on foreign tech, effectively helping a competitor in a strategic race.
Despite the U.S. easing export controls, China's government may restrict imports of NVIDIA's advanced chips. Beijing is prioritizing its long-term goal of semiconductor self-sufficiency, which requires creating a protected market for domestic firms like Huawei, even if Chinese tech companies prefer superior foreign hardware.
When NVIDIA CEO Jensen Huang praises Donald Trump's 'pro-energy' stance, the subtext is a strategic appeal. He is lobbying for the freedom to sell high-performance GPUs to China, despite significant national security implications recognized by the Defense Department.
Seemingly strange deals, like NVIDIA investing in companies that then buy its GPUs, serve a deep strategic purpose. It's not just financial engineering; it's a way to forge co-dependent alliances, secure its central role in the ecosystem, and effectively anoint winners in the AI arms race.
While NVIDIA laments lost revenue from export controls, those same policies blocked its primary Chinese competitor, Huawei, from accessing TSMC's advanced manufacturing. This prevented Huawei from launching a competing 7nm GPU, preserving NVIDIA's market dominance in China.
The US government's reversal on Nvidia H200 chip sales to China, now with a 25% tax, indicates a strategic shift. The policy is no longer a complete blockade but aims to keep China one generation of chips behind while generating significant tax revenue for the US.
A complete ban on selling chips to China is counterproductive. The ideal policy allows NVIDIA to sell chips that are one or two generations behind state-of-the-art. This strategy keeps Chinese firms dependent on the NVIDIA ecosystem, funds U.S. R&D with sales revenue, and hinders domestic competitors like Huawei from flourishing.
