Allowing American Companies to Export Lower-Grade Chips to China is Strategically Wise Decision
In a move that has sparked debate among U.S. policymakers, the U.S. government has recently allowed Nvidia and AMD to resume sales of their H20 and MI308 AI chips to China. The decision comes as both companies have designed these chips to comply with U.S. export controls, contrary to claims that they were developed to circumvent them.
The H20 and MI308 are second-tier chips with deliberately capped performance to meet export restrictions. However, some policymakers worry that any AI hardware export could aid Chinese military or surveillance applications, potentially threatening U.S. national security. Critics also point to an uneven playing field sentiment, where some U.S. businesses, academics, and smaller AI developers reportedly struggle to secure AI chips while large firms sell to China.
Sen. Elizabeth Warren (D-MA) condemned the decision, stating that it prioritizes sending these chips to China over domestic AI developers who face shortages. Rep. John Moolenaar (R-MI) also opposes the decision, citing strategic concerns.
However, proponents argue that blocking these exports punishes American companies following export rules, risks missing out on a large market (China accounts for about 13% of Nvidia’s revenue), and could allow Chinese companies to develop alternatives using less advanced chips and software ingenuity.
The opposition is less about the technical aspects of the chips—which are intentionally limited in capability—and more about the strategic and economic risks of enabling China’s AI advancement through even limited AI chip exports. The critics' concerns extend beyond technical capabilities; they reflect broader geopolitical concerns about China’s rising technological competitiveness and misuse of advanced U.S. technologies.
It's important to note that Chinese data centers are currently using only about 30 percent of their available compute capacity, indicating that China is not facing a shortage of computing power. The decision to allow the resumption of sales applies to AMD as well as Nvidia.
The decision to allow the sales could significantly boost China's AI development, potentially funding American R&D for the next generation of U.S. chips. Nvidia expects to earn $15 billion in revenue from China in the second half of 2025, which could be used to fund American innovation.
On the other hand, denying access to lower-tier U.S. chips like the H20 may raise costs for China at the margins, but it will not stop its AI ambitions, especially as Chinese firms accelerate efforts to build domestic alternatives. Blocking these chips would have created openings for Chinese competitors in the global AI race.
Overly broad export controls risk long-term strategic losses by pushing Chinese firms away from the U.S. AI ecosystem and encouraging the development and adoption of alternatives. Restricting exports not just hurts U.S. firms, it weakens America’s edge in AI and semiconductors and risks handing market share to foreign competitors.
This decision underscores the complexities and challenges of balancing national security concerns with the need for U.S. companies to compete globally and remain central to the technologies that will shape the future.
- The ongoing debate among U.S. policymakers concerns the resumption of sales of Nvidia and AMD's H20 and MI308 AI chips to China, as some worry that any AI hardware export could aid Chinese military or surveillance applications.
- Sen. Elizabeth Warren (D-MA) and Rep. John Moolenaar (R-MI) oppose the decision, expressing concerns about prioritizing exports to China over domestic AI developers who face shortages.
- Proponents argue that blocking these exports risks missing out on a large market, such as China, which accounts for about 13% of Nvidia’s revenue, and could allow Chinese companies to develop alternatives using less advanced chips and software.
- This decision could significantly boost China's AI development, potentially funding American R&D for the next generation of U.S. chips, as Nvidia expects to earn $15 billion in revenue from China in the second half of 2025, which could be used to fund American innovation.
