On Oct. 7, 2022, the Biden administration issued aggressive new regulations, aimed at making it harder for China to access and build high-end semiconductors.
Monday saw the publication of “Chip War,” a new book on global fights over semiconductors, written by Chris Miller, an associate professor of history at Tufts University’s Fletcher School and a visiting fellow at the American Enterprise Institute. I interviewed Chris by email over what the new developments mean and how they reflect a longer history of conflict. The interview has been lightly edited for style.
Q: Most semiconductor manufacturing happens outside the United States. So how is the Biden administration able to set limits on global production of semiconductors?
A: Making chips requires using ultracomplex machine tools, many of which are produced by a small number of firms in just a couple of countries. Certain of the most advanced tools are only produced by U.S. firms. The U.S. Department of Commerce can use a regulation called the “foreign direct product rule,” to restrict the use of these American-manufactured tools for making certain chips. Because Taiwanese, South Korean, and all other advanced chipmakers need these tools, the regulation is effectively global, so long as the United States retains a monopoly on these tools.
Q: Why did semiconductor production move out of the United States to countries like Taiwan in the first place?
A: Partly because it was cheaper. Differences in labor costs are a comparatively small part of the story. Cheap capital and government incentives were probably more important in explaining why it was less expensive to manufacture in East Asia than in the United States. However, cost is only part of the story. The other factor is that chipmaking involves huge economies of scale.
The company that has best capitalized on these economies of scale is the Taiwan Semiconductor Manufacturing Company (TSMC), which today produces more processor chips than any other company. Its vast scale not only provides cost efficiency but also allows TSMC to hone its production processes over a larger volume of chips. Today, TSMC is the world’s most advanced chipmaker, and almost all of its chips are made in a handful of massive factories in Taiwan.
Q: You explain in the book how Chinese President Xi Jinping sees semiconductors as a “vital gate” that can be used to control the Chinese economy. Why does China depend so much on foreign semiconductors?
A: For the past decade, China has spent more money importing chips than it has importing oil, because China has been unable to produce advanced chips like those in phones, PCs, and servers at home. No Chinese firm has reached the leading edge in chip manufacturing, and most foreign chipmakers have kept their advanced technology in their home countries. China therefore relies on foreign firms for advanced processor and memory chips.
The fact that China spends so much money buying chips has economic consequences. Chinese political leaders, though, may care more about the political ramifications, because China buys chips largely from firms from Korea, Taiwan, or the United States. This gives the United States leverage to control China’s access to chips for certain uses. Washington has used this leverage multiple times in recent years. For example, the United States has long imposed limits on sales of certain chips to Chinese military firms. The United States sought to limit the 5G telecoms business of Huawei by cutting the company’s access to certain types of chips. Now, via the rules announced this week, the United States is imposing limits on chips used for AI applications in data centers.
Q: Based on current manufacturing technologies and the experience of other countries, can China defy U.S. controls by creating its own native semiconductor industry?
A: At some point, China will likely catch up to today’s cutting edge, but it will be very expensive and time consuming. So long as existing U.S. controls remain in place, China will need to find alternative sources of tools and software if it wants to advance.
Today, China lags behind in most parts of the chip supply chain. Its companies use foreign, largely U.S. software to design chips, though Chinese firms are reportedly working on domestic replacements. China buys most of the tools and materials that it uses to fabricate chips from abroad, and even then, its companies are less effective at producing chips that competitors in Taiwan, South Korea or the United States.
China has spent many tens of billions of dollars since 2014, when it launched its current drive to build up its domestic chip industry, but it is still highly dependent on foreign technology. If China continues to spend at this rate, it will continue to make progress. But it will not be easy or quick because the tools China must try to produce domestically include some of the most complex and precise machinery in human history. For example, just one of the key parts of an extreme ultraviolet lithography system, the laser, requires more than 457,320 components.
Even if China has weaned itself of key foreign dependencies and created domestic versions of all the chips and chipmaking tools it imports today a decade from now, it may still be behind. The cutting edge of the industry will have moved forward. So China’s challenge is not only to replicate the current level of technology but to catch a train that is racing forward. It is far more challenging and expensive than catching up in any other sector of the economy.