Illustration of a giant hand with a USA cufflink stopping a delivery man carrying a box with the Chinese flag on it. — Illustration by iStock; Security Management

The Semiconductor Conundrum

By James Lewis

1 February 2023

Security Technology, February 2023

The United States rightly identified semiconductors as a key strategic technology in the competition with China. Semiconductors, commonly known as chips, are the heart of both modern weapons and commercial products. In January 2022, the Biden administration announced a major initiative to slow China and bring chip manufacturing back to the United States.

China has tried for decades to build a world-class chip industry, so far with limited success. For China to develop cutting edge chip technology, it needs outside support—particularly from the United States and Japan, whose companies build three-quarters of the advanced manufacturing equipment needed to make these high-tech products.

More importantly, allowing Chinese scientists and technicians to work in American chip companies has been a major source of capacity building for China, providing the know-how and skills needed for successful chip making. American venture capital firms were investing in and advising Chinese chip makers. Korean and Taiwanese firms had close links as well, and Taiwan was a source of executive and technical talent. Cutting these ties in October 2022 to stop China’s acquisition of American chip technology was the right strategic choice.

The new American restrictions impose export controls to block China’s ability to purchase and manufacture high-end chips, limiting the country’s ability to obtain advanced computing chips, develop and maintain supercomputers, and manufacture advanced supercomputers. The restrictions will significantly slow China’s chip efforts and will hold back Chinese semiconductor capabilities for years, but these effects will not be immediate, comprehensive, or ultimately permanent.

The United States and other Western countries spent 40 years putting China into their supply chains; taking it out cannot be done easily or quickly. Western companies rely on Chinese companies for their downstream supply chains (testing and package, for example) and it is not easy to replace them, particularly for the less sophisticated chips used in cars, appliances, and microelectronics—older, low-cost components like transistors or capacitators that are found in every electronic device. It made economic sense to move the production of these chips to China when China was not a national security threat to the United States and its Asian allies.

China is also a giant market, consuming 40 percent of global chip output (although many of these chips end up being assembled into devices for re-export), and was the biggest single market for semiconductor manufacturing equipment (SME) before the new rules were put in place. Companies are reluctant to give it up. China is already deeply embedded into the global chip supply chain with American, European, and Asian companies, and this close interrelationship is the source of friction over the new rules.

China has spent tens of billions of dollars to build its own chip industry and it will not simply give up. Chinese chip makers like SMIC and YMTC were positioned to compete in the global market before U.S. restrictions went into effect. China’s intentions in this space were similar to those seen with Huawei—to use government support and predatory trade practices to create a dominant position in the global semiconductor market. The new rules slow, but do not halt this trajectory.

What may stop China is its own policies. If a country spends billions of dollars for decades to build a chip industry without success (compared to other entrants who made similar efforts, like Taiwan and South Korea), we need to ask if there is some systematic problem that is holding it back. Some of this problem is related to corruption, with five senior executives linked to the government’s chip investment fund put under investigation last year. Some of this problem is waste, as state entities build “Potemkin fabs” mainly to score points with the Chinese Communist Party leadership. President Xi Jinping’s reversion to more state-controlled economic policies will also not help.

Export controls were a useful 20th Century tool for keeping advanced technology from the Soviet Union during the Cold War. But there was a clear bifurcation between opponents in the 20th Century, an Iron Curtain with very little trade between East and West. That is not the case with China. Export controls will be much harder to enforce and easier to circumvent in a globally connected economy. There was also, at least until the final decade of the Cold War, a strong degree of unanimity among Allies to keep military-related technology from the Soviets. This will be harder (but not impossible) to duplicate in the contest with China.

To be effective, any new arrangement to restrict access to semiconductor technology needs the participation of Japan, Korea, Taiwan, the Netherlands, and Germany (or the European Union), and perhaps Singapore, as well as the United States. Cooperation is not assured.

Japan is supportive, and given that the United States and Japan dominate the SME market, this is an invaluable advantage. But European governments are less committed to “decoupling.” The Netherlands company ASML provides irreplaceable technology. Taiwan, Korea, and Singapore are essential suppliers of chips and would like to keep selling. The United States can slow China’s ability to make its own advanced chips, but it will be harder to deny it access to chips, which are produced in the millions and sold both directly and through brokers, diversion, or smuggling. This is a complex, diplomatic landscape where the United States has signific support for restrictions, but also faces a degree of reluctance from some key partners.

The CHIPS Act, recently signed into law to support research and provide incentives to build semiconductor fabrication plants (fabs) in the United States, changes this competitive landscape for the better. It broke the long-standing taboo on industrial policy in the United States and it has incentivized allies to seriously discuss how to deal with China’s chip challenge. It restores funding for crucial R&D. Moving China completely out of the supply chain is not an achievable goal in the next few years, but shrinking its role in the global chip supply chain by using CHIPS Act subsidies and new restrictions on technology transfer is achievable and will improve the security of the United States and its Allies

The American tech and innovation base is recovering from the 1990s decision to cash in the “Peace Dividend” by shrinking federal government investment in research and technology. This decision may have made sense at the time, but not by 2015 when it became clear that China intended to replace a rule-based international order with one better suited to its needs—and that technology and innovation would be as important as military strength.

A new competition has begun, with the democracies again on the defensive. Technology is a key part of this contest. Unlike the Cold War against a slow-moving opponent in a bifurcated environment, the new approach will need to accommodate a degree of connection to China. This means that what we are seeing now with new restrictions, new subsidies, and, perhaps, new alliances, is only the start of what will be a long contest with chips at the center.

James Lewis, senior vice president and director of the Strategic Technologies Program at the Center for Strategic and International Studies, writes from Washington on technology and public policy. He has extensive politico-military, regulatory, and negotiating experience. Lewis has authored numerous publications and is quoted frequently in the media.