The Treacherous Silicon Triangle

Two of the biggest areas of risk in an increasingly fraught U.S.-Chinese relationship are the security of Taiwan and the security of the semiconductor supply chain. Each is high stakes and difficult in its own right. But taken together, they become even more challenging. Although each is critically important to the United States, their solutions do not always neatly align. The most obvious ways to reduce the risk in one area can easily increase risk in the other, and vice versa. Finding a way to manage this treacherous “silicon triangle” among Beijing, Taiwan, and Washington is thus one of the most important—and trickiest—challenges for U.S. foreign policy today.

The United States remains a world leader in semiconductor research and design, but its share of global manufacturing has fallen from 37 percent in 1999 to 12 percent today. Taiwan now accounts for the largest share of fabrication by far—producing 60 percent of the world’s chips and more than 90 percent of its leading-edge logic chips, key components in the world’s most advanced communications tools and computers and critical in the race for leadership in artificial intelligence. After Taiwan, the other top manufacturers of semiconductors are South Korea (which leads in the production of memory chips), Japan, and China, in that order. The United States has fallen to fifth place.

Policymakers broadly recognize the dangers of leaving the supply chain of such an essential component in an increasingly digital economy vulnerable to prolonged disruption—or worse, to deliberate denial by an adversary. The “chip famine” that emerged globally in 2020 wreaked havoc across a wide variety of industries. Worse still would be a chip shortage, or an embargo imposed by a hostile power, that crippled the production and maintenance of advanced U.S. weapons systems.

That fear is one of many reasons for concern about the security of Taiwan. Beijing’s escalating military and geopolitical pressure raises an enormous risk for the United States and its allies: if Beijing was able to successfully seize Taiwan, the Chinese leader Xi Jinping’s regime could suddenly gain dominance over the most critical manufactured commodity in the world—if the conflict did not disable or destroy much of Taiwan’s capacity to produce semiconductors.

Some strategists assume that the semiconductor industry constitutes a “silicon shield” for Taiwan, because chips are now so critically important to the global economy and to China’s economy that, in the absence of an extreme crisis or provocation, Chinese leaders would be unlikely to risk a conflict that could destroy or severely disrupt China’s own (and the world’s) supply of chips. What is more, the reliance of much of the rest of the world on Taiwan’s semiconductor industry gives a host of other governments an added stake in deterring conflict over Taiwan. Yet placing too much stock in this logic would be unwise; if Beijing decides to use force, it will be impelled principally by other political and geopolitical reasons.

In this silicon triangle, the United States has two intertwined interests: to safeguard the security of global supply chains for semiconductors—which must include some prudent degree of reshoring of production—and to ensure security and autonomous choice for Taiwan. The challenge is to forge a cooperative strategy in which the pursuit of each goal does not undermine the other. That requires building on the unique geopolitical strength of the United States—its dense web of partnerships and alliances—to enhance the resilience of the supply chain while at the same time working with Taiwan to jointly strengthen military and economic capabilities to deter aggression by Beijing. The United States needs investment from Taiwan to expand semiconductor manufacturing on American soil, thus creating jobs and enhancing the resilience of U.S. supply chains; Taiwan needs the security assistance of the United States to protect both its semiconductor industry and its democracy from aggression. Pursued together, while also deepening cooperation with other U.S. partners and allies crucial to global semiconductor production, these aims can reinforce one another in ways that will enhance both supply chain resilience and the security of Taiwan.

THE NEW OIL

In strategic terms, chips may be “the new oil,” as one formulation posits, but their journey from raw material to end use is far more complex. Their production depends on advanced designs and enormously sophisticated (and expensive) equipment. The most advanced chips are “fabbed” by state-of-the-art machines that use extreme ultraviolet lithography. These machines are produced by just one company, Advanced Semiconductor Materials Lithography, which is based in the Netherlands. In addition to raw materials and highly capital-intensive plants and equipment, production requires a close-knit, highly educated, and well-trained workforce of engineers and technicians. Once fabricated, chips must go through assembly, testing, and packaging, which are most often done in other plants based in other countries. And each fab is also dependent on continual program upgrades and technical maintenance. Often, these critical roles are divided among different countries. The United States, accordingly, must aim to ensure that the bulk of its demand for semiconductors (including the most advanced chips) is filled at each step in the supply chain by friendly countries committed to maintaining stable trading partnerships.

It is wise to seek to locate production of a wide variety of chips on U.S. soil by providing financial incentives for reshoring, as the 2022 CHIPS and Science Act does. But even if the United States doubles its share of global chip production in the coming years, it will still depend heavily on global supply chains, which must engage trusted partners. Principally, these will be the United States’ friends and allies, not just Taiwan but also France, Germany, Japan, the Netherlands, Singapore, and South Korea. (India is also poised to become a player in the industry, and the United States should help encourage investment in manufacturing there.) And even chip manufacturing in the United States will depend on working with the most technologically capable companies, many of them non-American. Taiwan Semiconductor Manufacturing Company, for example, is already building a $12 billion manufacturing plant for leading-edge chips in Phoenix. Last December, it announced it would build a second, even more advanced plant there, bringing TSMC’s total investment in the United States to $40 billion, already exceeding the $39 billion in subsidies for U.S. chip manufacturing provided by the CHIPS Act.

Even if the United States doubles its share of global chip production, it will still depend heavily on global supply chains.

In addition to subsidies, the United States must provide lower costs, ample infrastructure, expanded services, and engineering talent to attract further private investment in semiconductor manufacturing. Congress can help by extending 100 percent tax depreciation for short-lived capital assets (a rule that lapsed at the end of last year). Doing so would reduce the massive upfront costs of semiconductor manufacturing equipment needed to set up a new fab and offset other construction costs that TSMC has estimated to be four times higher in the United States than in Taiwan. Congress should also extend the chip manufacturing tax credit in the CHIPS Act beyond its 2027 sunset provision and broaden the credit to cover key material inputs and the manufacture of equipment. Federal and state government should also ease regulatory burdens to make it possible to construct plants in the United States more quickly. Given the industry’s relatively short technology cycles, multiyear environmental reviews will make a significant expansion in chip manufacturing a futile task. States can also help lure investment by ensuring adequate water and electricity supplies and providing incentives for related service and equipment companies, fostering the kinds of geographic clusters that helped drive Taiwan’s semiconductor miracle.

Partnering with Taiwan also offers huge opportunities on the technological front. U.S. research centers and universities can benefit from Taiwanese support on talent development, while Taiwanese firms can benefit from expanding research and development efforts in the United States. U.S. policy could help incentivize such collaborations. It could invite leading semiconductor companies from Taiwan (as well as from South Korea) to join the United States’ new public-private National Semiconductor Technology Center, while also building on efforts such as a collaboration between Purdue University and the Taiwanese chip designer MediaTek to develop a new joint chip design center. Policymakers can also enhance education and training in the United States, by encouraging the “semiconductor colleges” embedded in Taiwan’s top universities to partner with a proposed American Semiconductor Academy, as well as providing more funding for Taiwanese students to study in the United States and for American students to study in Taiwan. Broader economic and technological ties between the United States and Taiwan would also be strengthened by a treaty precluding double taxation of expatriate workers and by completing negotiations on a free trade agreement.

Yet as it fosters chip manufacturing at home, Washington must do more to deter a Chinese attack on Taiwan. Here, key lessons of the war in Ukraine are instructive. Although it is always better to deter an attack than to try to repel it after it has begun, both tasks depend on the delivery of effective weaponry—especially the kind of mobile weaponry that can help turn Taiwan into a “porcupine” that the People’s Liberation Army would be unable to swallow. Expertise and investment from Taiwan are essential to developing additional chip manufacturing capacity on U.S. soil; similarly, the United States must do more to help Taiwan get the additional advanced weapons it needs. Given the long delays in Pentagon procurement, Washington should pursue licensing agreements with Taiwan’s manufacturing sector to rapidly scale up local production of weapons such as Javelin antitank missiles, Stinger surface-to-air missiles, drones, and satellite communications systems.

STRONGER TOGETHER

As it strengthens partnerships with and attracts investment from friends and allies, the United States must also exercise vigilance about China’s semiconductor ambitions. It is not realistic or desirable to freeze China out of the global supply chain entirely. Instead, the goal must be to ensure that neither China nor any other potential future adversary can weaponize its position in semiconductor supply chains. On the domestic front, this will require more vigorous and transparent review of inbound investments by the Committee on Foreign Investment in the United States to ensure that potential adversaries do not acquire effective control over key U.S. technologies. There may also be value in reviewing and restricting outbound investments in critical foreign technologies, and in implementing new technology export controls to protect the most sensitive U.S. intellectual property—building on the Biden administration’s decision to restrict the export of technologies and tools that would help China make advanced logic chips. Another risk to guard against is Chinese dumping of certain kinds of lower-end chips, which could allow Beijing to drive out competitors from and achieve a dominant position in important segments of the market. These “legacy” chips are heavily used not only in consumer products but also in U.S. weapons systems. Chinese dominance of this market would thus pose serious economic and security risks.

According to U.S. intelligence and other analysts, Xi has set 2027 as the year by which China must be militarily ready to attack Taiwan. Although Xi may have been given second thoughts by Russian President Vladimir Putin’s disastrous invasion of Ukraine, there is still little time to lose in projecting U.S. readiness and resolve and in strengthening Taiwan’s ability to protect its democracy and the world’s microchip supply chain. Economic, technological, and strategic competition between China and the United States will remain the dominant feature of geopolitics for years, if not decades, to come. To enhance its chances of prevailing in this competition, the United States will need reliable international partners with whom it can reconfigure and strengthen its semiconductor supply chain. No partner is more important in this effort than Taiwan.