The Chinese People’s Liberation Army (PLA) is using US-designed artificial intelligence chips as part of its upgrade and modernization program but an embargo could be difficult or dangerous to enforce, analysts say.
A report prepared by the Center for Strategic and Emerging Technology (CSET) notes that GPUs, FPGAs and some application-specific circuits can be used for AI. It analysed thousands of public contracts awarded by PLA units and state-owned defense enterprises in 2020 and notes that the AI chips identified in the purchase records nearly all were designed by Nvidia, Xilinx (now AMD), Intel or Microsemi – and manufactured in South Korea or Taiwan.
There were no records of orders for high-end AI chips designed by Chinese companies. However, it is also noted that the data was obtained from public records and there may be classified records of purchases with a different balance.
The report considers two ways of limiting Chinese military access to AI chips: one is a ban on sales to intermediary Chinese military suppliers. The second an embargo the export of AI chips to China. “However, both of these options have serious limitations, and could prove counterproductive to US national and economic security interests,” the authors conclude.
The report notes that the Chinese market accounts for 25 percent of global AI chip consumption, with AI chips sold to China in 2021 amounting to an estimated value between $2.5 billion and $5 billion. AMD, Nvidia and Intel derived 25, 26 and 27 percent of net revenue from China in 2021.
Therefore, the authors conclude, that adopting an embargo on AI chip exports suddenly “could have severe and potentially catastrophic consequences for the US semiconductor industry and long-term technological innovation.”
Related links and articles: