Failures, Successes, and Challenges in Advanced Semiconductor Manufacturing: Lessons from the Soviet Union, China, and Taiwan

AdobeStock 452152099 1 scaled

Semiconductors are the technological backbone of modern electronics and, by extension, the global economy. The policies governing their production and distribution significantly impact economic growth and technological advancement. As the US is attempting to reshore advanced semiconductor manufacturing, it is critical to understand the failures of the Soviet and Chinese semiconductor manufacturing programs and the successes and challenges facing Taiwan’s advanced semiconductor manufacturing sector.

The US does not seem on track to reshore advanced semiconductor manufacturing. While the US is a major player in the semiconductor design space, it has largely ceded manufacturing, assembly, testing, packaging, and other elements of the supply chain elsewhere. Japan, South Korea, and the Netherlands play vital parts in the semiconductor supply chain, but Taiwan is the undisputed leader in manufacturing. Over the past few decades, strategic Taiwanese industrial policy has helped it develop into a semiconductor manufacturing powerhouse, producing over 60% of global semiconductors1 with its leading semiconductor company, TSMC, accounting for 92% of the world’s most advanced semiconductors.2 Through the 2022 CHIPS and Science Act, the US made a significant attempt to reshore advanced semiconductor manufacturing, but since initial 1980s initiatives like SEMATECH, US industrial policy has not effectively incentivized companies to manufacture advanced semiconductors domestically. The risks of a lagging domestic semiconductor industry reverberate in cautionary tales from the Soviet Union’s failed industrial espionage initiatives and its impact in Russia today. At the same time, the risks of massive state-sponsored semiconductor initiatives are evident in China’s semiconductor investments mishandling talent, overregulating, and corruptly investing in Ponzi schemes. Today, TSMC is still making the most advanced semiconductors in Taiwan,3 and its subsidized Arizona fab is not only behind schedule and over budget but also already last-generation,4 and DC-groupthink mirror imaging US security concerns on Taiwan may cause the CHIPS and Science Act to follow in the footsteps of similar failed Soviet and Chinese initiatives. Simultaneously, however, rosy perceptions of Taiwan’s ostensibly positive semiconductor development gloss over the industry’s unique challenges in Taiwan, which presents an opportunity for alternative analysis. Addressing these challenges in Taiwan can provide a strategically and economically viable direction for semiconductor investments in which the US has a comparative advantage.

Soviet reliance on industrial espionage for semiconductor development left Russia technologically disadvantaged in military development

Soviet semiconductor development was fundamentally centered around industrial espionage instead of innovation. As the Soviet Union realized the strategic nature of semiconductors in the 1950s, Soviet scientists and spies began infiltrating US semiconductor companies and research centers.5 In 1958, Khrushchev even built an entire city, Zelenograd, around semiconductor production.6 A year later, the CIA assessed that Soviet semiconductor technology was often crude copies of Western designs and remained 5-7 years behind the US.7 By the 1970s, Soviet leaders provided Zelenograd’s scientists even less leeway to experiment, implementing blanket directives to abandon original development and copy American semiconductor designs instead.8 Around the same time, Intel co-founder Robert Noyce realized that focusing on the consumer market rather than the US military would accelerate semiconductor demand and innovation to new heights,9 but the CIA found that Zelenograd’s dependence on the Soviet military as its primary customer only grew.10 Nevertheless, the Soviet semiconductor industry would perpetually remain 5-7 years behind the US until the Soviet military was dissolved in 1992, after which Zelenograd’s uncompetitive semiconductor fabs shut down.11 Even worse than the Soviet’s perpetual technology lag, however, the Soviet pirate and copy strategy allowed the CIA to waste Soviet resources by purposefully designing faulty computer chips that passed Soviet quality tests but failed in operation. The Soviet pirate and copy strategy was also downright dangerous, as the CIA was able to cause the “most monumental non-nuclear explosion and fire ever seen from space” by shipping Trojan horse kill switches in the KGB Directorate T’s pirated microchips, resulting in a 3-kiloton explosion destroying a Siberian natural gas pipeline.12

Russia’s loss of domestic semiconductor production has left Russia’s civilian and military industries at a technological disadvantage. Due to Russia’s reliance on foreign semiconductor manufacturing mainly centered around US partners like Taiwan and South Korea, its industries have been forced to adopt low-tech designs to ensure at least some supply chain continuity if the US and its allies were to restrict chip sales. Thus, only 5% of the weapons Russia used in Syria were precision-guided.13 Even with such self-imposed restrictions, US sanctions after Russia invaded Ukraine still forced Russia to extract dishwasher and refrigerator semiconductors for military use.14 Overall, Russia’s non-existent semiconductor manufacturing base coupled with US sanctions have drastically reduced Russian economic and military might for the foreseeable future.15

China’s advanced semiconductor industry development failed due to talent drain, industrial espionage, overregulation, and corruption

Although the Chinese Communist Party realized the importance of semiconductor development in the 1950s, Mao Zedong’s radicalism drastically inhibited China’s semiconductor potential. As Mao killed and imprisoned university-educated engineering talent during the Cultural Revolution, he implemented impetuous policies declaring “all people must make semiconductors” during the Great Leap Forward.16 Thus, many talented Chinese university graduates fled to Taiwan, Hong Kong, Singapore, and California, with some, like Morris Chang, becoming kingpins of the semiconductor industry. By the time Deng Xiaoping took Mao’s place, China seemed hopelessly behind Taiwan, South Korea, Japan, and the US in semiconductor technology. Yet China still considered semiconductors strategically important and thus tried in vain to control chipmaking, inadvertently stifling innovation by shrouding the sector in bureaucracy. Therefore, China’s electronics industry remained almost entirely dependent on foreign chips from the US, Japan, and Taiwan for decades.17

As China entered a new era, semiconductor startups benefited from vast government support, but much of it was misguided. Zhang Rujing—China-born, Taiwan-educated, and US-trained—founded SMIC, China’s largest pure-play semiconductor foundry, in 2000.18 Partially state-owned, SMIC tried replicating Taiwan’s semiconductor success through TSMC’s playbook. SMIC extensively poached Taiwanese, South Korean, and American semiconductor experts and received state subsidies, corporate tax holidays, and reduced tax on Chinese sales.19 China also covertly supported its semiconductor industry through cyber espionage.20 Still, SMIC and other Chinese fabs remain at least 5 years behind TSMC’s leading semiconductor technologies.21 During Xi Jinping’s tenure, he launched the China Integrated Circuit Industry Investment Fund, or “Big Fund,” intending to make China’s semiconductor industry self-sufficient by 2025. While some of its investments in companies like SMIC have been relatively successful in helping China develop trailing-edge semiconductor production capabilities, the “Big Fund” has also been submerged in corruption scandals and Ponzi schemes with its investments in now-bankrupt companies like Tsinghua Unigroup. Due to such corruption, the “Big Fund” has experienced a massive brain drain, with 25% of its talent leaving since 2020.22

Although the most successful semiconductor manufacturing nation, Taiwan’s semiconductor industry faces significant challenges from geostrategic risk, talent, land, energy, water, and food

Taiwan took a different route than the Soviet Union and China, seeking to innovate instead of industrial replication. Government support, foreign investment, STEM education, entrepreneurship, and industry-academia collaboration allowed Taiwan to become an advanced semiconductor manufacturing powerhouse. Throughout the 1980s, authoritarian government directives spearheaded by Taiwan’s economic minister KT Li allowed Taiwan to develop a semiconductor industry cluster in Hsinchu Science Park. Taiwan’s central government recognized that Taiwan needed to create an indigenous innovative electronics industry as cheap labor shifted to China, or else get caught in the middle-income trap. As Taiwan pursued semiconductor technology transfers from US companies in the 1970s, Taiwan’s government simultaneously offered incentives to overseas Taiwanese like Morris Chang to start businesses in Taiwan. Chang soon pioneered TSMC’s dedicated foundry business model in Taiwan’s Industrial Technology Research Institute (ITRI), receiving initial funding from the Taiwanese government and Philips, which owned ASML—the world’s most advanced lithography company. Even though Taiwan’s government and Philips sold off most of their shares, TSMC’s first-mover advantage in the dedicated foundry market and collaboration with academic research centers like ITRI, National Tsing Hua University, and National Yang Ming Chiao Tung University allowed it to couple narrowly focused industrial R&D with highly-educated engineering talent. Such an arrangement permitted Hsinchu to develop similarly to Silicon Valley into the highest average income area in Taiwan, with Stanford and Berkeley also synergistically feeding top engineering talent to Silicon Valley companies. Simultaneously, Taiwan’s government was also incubating a comprehensive semiconductor ecosystem in Taiwan’s science parks, symbiotically integrating low-cost design, assembly, testing, and packaging with advanced manufacturing.23

Although Taiwan’s semiconductor industry is remarkably successful, primary-source on-site interviews conducted in Taiwan with government officials, industry experts, academic researchers, and local residents have revealed the challenges that the industry faces in Taiwan. First, Taiwan faces pressing geostrategic risks from China and natural disasters. Moreover, it also faces a talent drain due to low wages and racism. Taiwan also has minimal land area, causing tensions with local residents over industrial expansion. Furthermore, Taiwan suffers from energy security risks in power production and infrastructure. Taiwan has already faced major water shortages, a resource critical for semiconductor manufacturing, which climate change will only exacerbate. Lastly, Taiwan’s limited land facilitates food security threats as environmental contamination destroys the sparsely available farmland. The following paragraphs will also highlight the demographic groups left behind by Taiwan’s heavy-handed semiconductor industry favoritism from Taiwan’s martial law era until now.

Perhaps most concerning for US national security, Taiwan’s semiconductor industry is uniquely exposed to geostrategic risks. Beyond the fact that unifying Taiwan is China’s main strategic priority, much of Taiwan’s advanced semiconductor industry is geographically concentrated in Hsinchu, which sits atop an equatorial seismic fault line. Thus, Taiwan’s semiconductor supply chain is uniquely vulnerable to sabotage, surprise attacks, earthquakes, and typhoons. Taiwan’s small size also makes its semiconductor industry fundamentally dependent on foreign governments and clients, who may impose their demands on Taiwanese semiconductor companies like TSMC. Yet, TSMC does not plan on building the world’s most advanced semiconductor manufacturing fabs outside of Taiwan. To US defense planners, such dependence creates a paradox in which it is necessary to defend Taiwan to keep the US economy afloat and ensure national security, but the only way of defending Taiwan is by maintaining weapons systems dependent on a steady advanced semiconductor supply from the island.

Another challenge Taiwan’s semiconductor industry suffers from is brain drain as talented youth seek opportunities elsewhere. First, Taiwan’s semiconductor industry often looks for unquestioning and obedient university-educated workers. Thus, top entrepreneurial talent and non-college-educated students are often left out of the sector Taiwan’s government favors the most. Beyond the industry’s specific target demographic, many of these Taiwanese youth are also emigrating due to a perceived lack of opportunities on the island, low salaries, and excessively high property costs.24 Even worse, Taiwan has one of the lowest fertility rates in the world at less than 1 birth per woman, and it is set to become the world’s lowest by 2025.25 Additionally, Taiwan’s wealth is quite centralized among elite business owners and management positions are often occupied by elderly industry vanguards, further limiting economic mobility and innovative opportunities for young Taiwanese. While Taiwan’s government has implemented a New Southbound Policy to encourage immigration from Southeast Asia, South Asia, and Oceania to fill semiconductor manufacturing jobs,26 its efforts are often thwarted by racism and discrimination from Taiwan’s mostly homogenous Han Chinese majority. As Taiwan and India were about to sign a memorandum of understanding to allow 100,000 Indian workers to work in Taiwan, several Taiwanese netizens protested against the decision, terming Indian men “rapists.” While Chinese propaganda is undoubtedly helping to push such narratives in Taiwan,27 some Taiwanese factories and a county government implemented restrictive discriminatory policies on Southeast Asian migrant workers during the COVID-19 pandemic in 2021.28 Beyond simple propaganda, China uses Taiwan’s discrimination against migrant workers, coupled with its economic leverage, as a psychological cudgel to boost its influence in ASEAN at the expense of Taiwan.29

Although often prioritized, Taiwan’s semiconductor industry is running out of the sparsely available land remaining in Taiwan. Behind Bangladesh, Taiwan is now the second most densely populated non-city-state country in the world.30 When Hsinchu Science Park was created in 1980, many local farmers were initially displaced, and low-wage laborers were eventually priced out of Hsinchu due to exponentially increased rent and property prices. As the Hsinchu semiconductor industry was developing, new fabs released dangerous pollutants into the air, affecting nearby residents in the crowded city.31 Lax government environmental regulations and a push for rapid economic development during Taiwan’s martial law era ensured that those negatively affected by semiconductor development would be silenced. While democratic Taiwan has taken steps to regulate environmental pollution, accidents still pose acute environmental hazards to local residents.32 Now, however, northern Taiwan’s Hsinchu Science Park has become so crowded with semiconductor fabs that there is no more space between Hsinchu City to the west and the Baoshan mountains to the east. Thus, semiconductor companies are diversifying to less-wealthy regions in central and southern Taiwan. Besides spreading the risk of environmental disasters and making land much more expensive for other industries and local residents, Taiwan’s advanced semiconductor industry continues to encroach upon farmland and traditional indigenous areas. Today, Taiwan’s semiconductor industry still wields significant influence over Taiwan’s government, and the demographic groups left behind are often sidelined.

Beyond land, Taiwan’s semiconductor industry faces critical energy security vulnerabilities. As Taiwan seeks to transition to renewable energy and phase out nuclear power, its energy infrastructure is becoming increasingly tenuous. In 2022, over 99% of Taiwan’s energy was imported, and Taiwan had 8 days of LNG stockpiled in 2 terminals. By 2027, Taiwan is seeking to build 3 more LNG terminals, increasing its LNG reserve stockpile to 14 days.33 Simultaneously, Taiwan seeks to increase solar and wind production,34 but Taiwan’s limited land makes such renewable initiatives challenging. Even if Taiwan successfully hits its LNG reserve and renewable targets, its energy supply would likely still be inadequate in a blockade or natural disaster scenario. Additionally, renewables introduce further variability into Taiwan’s already-stretched power grid, which will have larger and more frequent voltage changes.35 As evidenced in 2022, even a 90-minute major power outage in Taiwan could affect the semiconductor supply chain for months.36

Climate change and resulting water risks also pose major threats to Taiwan’s semiconductor industry. TSMC’s water-intensive advanced semiconductor fabs are increasingly built in southern Taiwan, which was hit hard by a drought in 2021, and the region’s water supply is only projected to become more tenuous with climate change. While TSMC was able to truck in water from northern and central Taiwan to supply its southern Taiwan fab, Taiwan’s rainfall is predicted to decrease by 10% overall by 2050, so such options may no longer be feasible. As Taiwan’s semiconductor industry continues to expand, southern Taiwan now has a water deficit of around 390,000 metric tons a day.37 Thus, Taiwan has been building desalination plants to aid in securing a stable water supply, but such plants are extremely energy-intensive, stressing Taiwan’s already overstretched energy infrastructure.38 Additionally, the southern Taiwan cities of Kaohsiung and Tainan are combining reservoirs and planning new water reclamation centers to feed TSMC’s growing southern presence,39 but these pose unique risks to the environment and local residents. Researchers at Kaohsiung’s National Sun Yat-sen University have noted how the semiconductor industry is a leading contributor to drinking water overuse and pollution, along with ocean contamination.40 Taiwan’s limited water supply often goes to semiconductor fabs through local pipelines or water trucks ahead of Taiwan’s residents, farmers, and other industries, which restricts the amount available for private consumption and irrigation, posing economic and food security risks. Thus, semiconductor fabs have been trying to increase recycled water consumption, but such in-house measures are again often expensive, energy-intensive, and imperfect in cleaning pollutants.41

Lastly, an excessive focus on growing Taiwan’s semiconductor industry harms Taiwan’s food security. About 70% of Taiwan’s food is imported, which will likely rise as a lack of land and water increasingly strangles Taiwan’s farmers. As Taiwan seeks to ensure a stable water supply for its semiconductor fabs, Taiwan’s government has prohibited rice farmers in southern Taiwan from planting their crops for the past three years. Thus, those farmers increasingly resort to desperate measures to earn money, such as converting farmland into solar panel farms to sell electricity to the semiconductor industry. However, these panels have negative environmental implications as they bleed toxic lead and cadmium minerals into farmland, contaminating future prospects of healthy crop growth. Therefore, Taiwan’s already limited food supply will be reduced even further with short- and long-term implications.42 With dwindling self-sufficient food supplies in a blockade or natural disaster scenario, semiconductor fab workers and other Taiwanese residents may face critical food shortages.

Soviet, Chinese, and Taiwanese failures, successes, and challenges in creating advanced semiconductor manufacturing industries pose important lessons for the US in minimizing industrial espionage, overregulation, and corruption while maximizing the strategic and economic rationale for ensuring abundant talent inflow and natural resources

While it is unlikely that the US will replicate Soviet industrial espionage techniques, significant lessons and implications can be drawn from the Soviet Union’s state-subsidized semiconductor industry failure. First, copying Taiwanese semiconductor designs through industrial espionage would inefficiently keep the US perpetually behind, damage US reputation, and set dangerous precedents facilitating potential sabotage and more intellectual property theft. Second, considering that most US military infrastructure is built around high-tech designs, losing access to Taiwan’s semiconductor manufacturing facilities would be even more devastating for the US military than US export controls on Russia’s military-industrial sector, considering Russia already planned its military with low-tech solutions in mind. Even if the US military integrates more low-tech solutions, semiconductor supply is bound to be affected since Taiwanese and Chinese companies produce 75% of the world’s legacy chips.43 If China were to cut off Taiwan’s semiconductor supply, defending Taiwan would be much more difficult. Beyond the military, losing access to semiconductors crippled Russia’s electronics and consumer industry, and losing TSMC would likely hit the US economy and technology-centered society even harder.

As the US considers certain semiconductor development pathways amidst massive federal semiconductor investment, China’s approach to building a domestic semiconductor industry offers cautionary tales in talent management, overregulation, and corruption that the US should avoid. Mao-era China’s anti-intellectualism stalled China’s semiconductor development for decades. Thus, the US ought to be wary of growing anti-science trends and increased nativism among the American populace and elected officials. Top engineering and supply chain management talent from around the world, including China, should be able to innovate freely in the US. While the danger of Chinese industrial espionage exists, banning top Chinese talent from the US semiconductor industry would do more harm than good, helping China reverse its own brain drain.44 On the flip side, Biden’s current policies banning US talent at Chinese chip firms seem quite effective at keeping American semiconductor workers and many Taiwanese semiconductor workers who hold dual nationality out of China, incentivizing them to go to Taiwan instead.45 Biden’s policies may not force American semiconductor workers to return to the US, but any such bureaucratic policies may be misguided anyway. In pursuit of domestic semiconductor production, China overregulated its semiconductor industry, which the US may be in danger of doing already. For legislation like the CHIPS and Science Act aimed at attracting investment in US semiconductor manufacturing, the US should put strategy over economic protectionism. Producing semiconductors on US soil is more important than regulating the percentage of US workers and parts within the fabs, and companies like TSMC ought to be able to bring non-unionized Taiwanese workers to build and staff the fabs if it means the US will have a higher global share of advanced semiconductor manufacturing. After all, Chinese state-sponsored companies like SMIC poached talent mainly from Taiwan and South Korea, allowing SMIC to become China’s most advanced semiconductor manufacturer. The US should welcome Taiwanese workers wishing to work in TSMC’s Arizona facility with the prospect of higher wages and a green card, which would further contribute to US strategic interests by safely harboring critical Taiwanese semiconductor talent. Beyond overregulation, China’s massive semiconductor investments also offer lessons in corruption, mismanagement, and exaggerated Ponzi schemes. The US ought to conservatively provide funds to established advanced semiconductor manufacturing titans like TSMC and Samsung based on their track record of success to avoid China’s pitfalls. Even if Intel is a US company, it has fallen behind in the advanced semiconductor manufacturing race, so the US should primarily pressure TSMC and Samsung to build more advanced semiconductor fabs in the US through more subsidies.

If the US can enhance its domestic competitiveness to capitalize on Taiwan’s challenges in terms of geostrategic risk, talent, land, energy, water, and food, it could likely better incentivize domestic advanced semiconductor manufacturing. China offers a major cautionary tale in excessively trying to replicate how Taiwan developed a successful semiconductor manufacturing industry by throwing state-sanctioned resources at the industry rather than evaluating the challenges that such an industry poses in Taiwan. Like China, the US has much more material resources than Taiwan; yet, US semiconductor policy leadership is at risk of making the same mistake as China.46

As beneficial as free trade and semiconductors are to Taiwan’s economy, keeping advanced semiconductor manufacturing in Taiwan poses an existential risk to Taiwan itself, the US, and the global economy. Even in Taiwan, some view TSMC’s success as a curse for putting Taiwan further in the danger zone.47 If anything were to happen to Taiwan’s semiconductor facilities, the US would be almost powerless to support Taiwan due to the resulting semiconductor shortage. Even if ASML were to quickly ramp up lithography equipment production and ship the lithography machines to US-based semiconductor fabs, ASML relies on the advanced chips that TSMC makes, posing a crucial chokepoint in a disaster or contingency scenario.48

Although bringing advanced semiconductor manufacturing to the US is critical, the US should tread carefully. While some have suggested overtly threatening TSMC with export controls to simultaneously roll out advanced technologies in the US and Taiwan or match its US and Taiwan investments 1:1, such actions would only feed China’s propaganda machine and turn Taiwanese people against the US. After all, China tried luring Taiwanese workers with massive paycheck carrots to build its indigenous semiconductor manufacturing capabilities, but such employees are often motivated by money over advancing technology. Forcing TSMC with a stick to build the most advanced semiconductor fabs in the US would presumably lower Taiwanese worker morale even further, and any such fabs would likely be unproductive. In contrast to the 1980s US-Japan semiconductor trade war, there is a decent likelihood that forcing Taiwan to give up its semiconductor technology would hand the world’s most advanced semiconductor technologies and “unsinkable aircraft carrier” right next door to the US’ primary geopolitical rival. Additionally, unilateral covert influence operations amplifying the concerns of anti-semiconductor demographic groups in Taiwan is an option, but such actions may sow distrust and aid Chinese anti-US propaganda if discovered. Instead, the US should seek to make building TSMC semiconductor fabs in the US as economically profitable as possible. TSMC generates most of its revenue from US companies, and the US could push private sector corporations and investors to pressure TSMC to diversify its advanced semiconductor fabs to the US. Furthermore, the US could optimize the CHIPS and Science Act and October 7th rules toward other Taiwanese semiconductor industry challenges. After all, Taiwanese policymakers seem to view the CHIPS and Science Act and October 7th rules as economically beneficial for TSMC and by extension, Taiwan.49

In terms of the Taiwanese semiconductor industry’s talent issue, the US should take crucial steps to encourage semiconductor workers to come to the US. While a similar wealth inequality problem exists in the US, Taiwanese workers are paid an average of $22,000 a year.50 Thus, the US should encourage talented Taiwanese engineers to come to the US by providing research opportunities, subsidizing higher salaries, emphasizing lower property costs, and offering paths to a green card. Furthermore, while discrimination undoubtedly exists in the US, Taiwan’s mostly homogenous ethnic composition makes it harder for Southeast and South Asian workers to avoid structural racism in Taiwan. Thus, by ensuring US industry is welcoming the best and brightest worldwide, the US can attract more skilled semiconductor engineers. Beyond immigration, however, the US ought to implement comprehensive educational reforms to nurture domestic STEM talent—simply investing more in STEM education through the CHIPS and Science Act is not enough.

Limited land and environmental concerns in Taiwan pose an opportunity for the US to attract advanced semiconductor investment. While US property rights and environmental regulations were stronger when Taiwan first built out its semiconductor industry, both countries are now democracies with similar rights and regulations. Furthermore, considering that the US has much more land, semiconductor priorities are less likely to conflict with those of local residents. Thus, the US should emphasize to semiconductor companies how its plentiful land makes buying land for new fabs cheaper upfront and ensures that those fabs are less likely to cause local residents to complain about pollution.

Given Taiwan’s energy insecurity, the US should emphasize American energy self-sufficiency to semiconductor companies. Considering how crucial stable power is for semiconductor fabs, the US ought to also create a resilient energy grid, especially in areas without resilient grids like Texas, where Samsung is building a new semiconductor fab. Besides providing resilient grids, the US should subsidize backup generators and voltage regulators in semiconductor fabs, thus guaranteeing stable power supplies.

Regarding water, Taiwan seems uniquely vulnerable, and the US should preemptively prepare for similar water challenges. TSMC’s new fab is in the deserts of Arizona, and the US needs to dramatically enhance the drought-prone West Coast’s water infrastructure to support such water-intensive semiconductor fabs. If the US can ensure a stable water supply, TSMC could build additional advanced fabs near its initial fab in the future. Creating an enhanced water infrastructure would also placate the concerns of local residents, farmers, and other industries in Arizona, paving the way for further population and industry growth in one of the fastest-growing places in the US.

Taiwan also faces significant food security challenges, which the US is well-positioned to alleviate. The US is the largest agricultural exporter in the world and will likely not face a food security crisis in a contingency scenario, which would plausibly be attractive to TSMC. Additionally, the US has sufficient land to easily separate farmland, local residents, and semiconductor fabs, ensuring that semiconductor contaminants do not affect food production and locals.

Conclusion

As evidenced by the Soviet Union, China, and Taiwan, the global semiconductor manufacturing industry is strategically critical but exceedingly complex. The Soviet Union’s reliance on industrial espionage, subsequent difficulties, and China’s challenges with talent management, overregulation, and corruption serve as cautionary tales, emphasizing the need for private-sector innovation. Despite Taiwan’s current status as a linchpin of semiconductor manufacturing, it faces significant geostrategic risks, talent retention issues, and natural resource limitations. In attempting to reshore advanced semiconductor manufacturing, the US ought to create an economically-rational atmosphere for semiconductor manufacturing, addressing talent retention, environmental sustainability, energy stability, and robust infrastructure. Such an approach would best strengthen national security and help the US maintain a resilient and advanced domestic semiconductor manufacturing industry.

References

Featured/Headline Image Caption and Citation: A fab worker holding a semiconductor chip | Image sourced from Maksim Shmeljov/Adobe Stock

  1. “Taiwan’s dominance of the chip industry makes it more important.” 2023. The Economist. https://www.economist.com/special-report/2023/03/06/taiwans-dominance-of-the-chip-industry-makes-it-more-important.
    ↩︎
  2. Cronin, Richard. 2022. “Semiconductors and Taiwan’s “Silicon Shield.”” Stimson Center. https://www.stimson.org/2022/semiconductors-and-taiwans-silicon-shield/.
    ↩︎
  3. Lee, Yimou, Ann Wang, and Stephen Coates. 2022. “TSMC starts volume production of most advanced chips in Taiwan.” Reuters. https://www.reuters.com/technology/tsmc-starts-volume-production-most-advanced-chips-taiwan-2022-12-29/.
    ↩︎
  4. Lovejoy, Ben, Rikka Altland, Laura Rosenberg, and Filipe Espósito. 2023. “TSMC Arizona chip plant will be a paperweight, says analyst.” 9to5Mac. https://9to5mac.com/2023/09/11/tsmc-arizona-chip-plant/.
    ↩︎
  5. Miller, Chris. 2022. Chip War: The Fight for the World’s Most Critical Technology. New York: Scribner.
    ↩︎
  6. Usdin, Steven T. 2008. “Zelenograd, the Soviet Silicon Valley, 1962–1965.” In Engineering Communism: How Two Americans Spied for Stalin and Founded the Soviet Silicon Valley, 203-225. New Haven: Yale University Press.
    ↩︎
  7. “Production of Semiconductor Devices in the USSR.” 1959. CIA. https://www.cia.gov/readingroom/docs/CIA-RDP79R01141A001500150002-6.pdf. ↩︎
  8. Ter, Aram. 2014. “Computers in the USSR: A story of missed opportunities.” Russia Beyond. https://www.rbth.com/science_and_tech/2014/09/24/computers_in_the_ussr_a_story_of_missed_opportunities_40073.html.
    ↩︎
  9. Berlin, Leslie. 2007. The Man Behind the Microchip: Robert Noyce and the Invention of Silicon Valley. New York: Oxford University Press, USA.
    ↩︎
  10. “USSR Seeks to Build Advanced Semiconductor Industry With Embargoed Western Machinery.” 1976. CIA. https://www.cia.gov/readingroom/docs/DOC_0000498593.pdf.
    ↩︎
  11. Prokhorov, Sergei P. 1999. “Computers in Russia: science, education, and industry.” IEEE Annals of the History of Computing 21, no. 3 (July-September): 4-15. 10.1109/85.778978.
    ↩︎
  12. Harada, Lawrence K. 2010. “Semiconductor Technology and US National Security.” US Army War College. https://apps.dtic.mil/sti/pdfs/ADA526581.pdf.
    ↩︎
  13. Jadallah, Ahmed. 2018. “What Kind of Victory for Russia in Syria?” Army University Press. https://www.armyupress.army.mil/Journals/Military-Review/Online-Exclusive/2018-OLE/Russia-in-Syria/.
    ↩︎
  14. Whalen, Jeanne. 2022. “Sanctions force Russia to use computer chips from refrigerators, dishwashers in military equipment, Gina Raimundo tells Senate.” Washington Post. https://www.washingtonpost.com/technology/2022/05/11/russia-sanctions-effect-military/.
    ↩︎
  15. Urusov, Pavel. 2023. “Vital Microchip Sanctions Will Hit Russian Computing Power Hard.” Carnegie Endowment for International Peace. https://carnegieendowment.org/politika/90250.
    ↩︎
  16. Lee, Keun. 2021. China’s Technological Leapfrogging and Economic Catch-up: A Schumpeterian Perspective. London: Oxford University Press.
    ↩︎
  17. Miller, Chris. 2022. Chip War: The Fight for the World’s Most Critical Technology. New York: Scribner.
    ↩︎
  18. Chafkin, Max, Sarah Frier, and Dana Hull. 2023. “Richard Chang: Taiwan’s Silicon Invasion.” Bloomberg.com. https://www.bloomberg.com/news/articles/2002-12-08/richard-chang-taiwans-silicon-invasion.
    ↩︎
  19. Y, Jon. 2022. “TSMC’s renegade genius.” Meet Global. https://meet-global.bnext.com.tw/articles/view/47800?
    ↩︎
  20. Antoniuk, Daryna. 2023. “China-based spies are hacking East Asian semiconductor companies, report says.” The Record by Recorded Future. https://therecord.media/china-budworm-apt27-east-asia-semiconductor-companies.
    ↩︎
  21. Wolff, Ilene, Alan Patterson, and Sally Ward. 2023. “SMIC Well on Its Way to 5-nm Breakthrough, Observers Say.” EE Times. https://www.eetimes.com/smic-well-on-its-way-to-5-nm-breakthrough-observers-say/.
    ↩︎
  22. Zhang, Erchi. 2022. “Graft Scandal Casts Long Shadow Over China’s Chipmaking Ambitions.” Caixin Global. https://www.caixinglobal.com/2022-08-08/cover-story-graft-scandal-casts-long-shadow-over-chinas-chipmaking-ambitions-101923281.html.
    ↩︎
  23. Based on interviews with senior officials at Hsinchu Science Park.
    ↩︎
  24. Ryan, Matthew. 2023. “Why Taiwan thinks it’s a poor loser when it’s actually a success.” CommonWealth Magazine. https://english.cw.com.tw/article/article.action?id=3491.
    ↩︎
  25. Jennings, Ralph. 2022. “‘Ticking time bomb’: Taiwan’s fertility rate to become world’s lowest by 2035.” South China Morning Post. https://www.scmp.com/economy/economic-indicators/article/3197997/taiwans-fertility-rate-set-become-worlds-lowest-2035-ticking-demographic-time-bomb-grows-louder.
    ↩︎
  26. Glaser, Bonnie S. 2018. “The New Southbound Policy.” Center for Strategic and International Studies. https://southbound.csis.org/.
    ↩︎
  27. Singh, Vikrant. 2023. “How racism overshadowed India-Taiwan co-operation.” WION. https://www.wionews.com/world/explained-how-racism-overshadowed-india-taiwan-co-operation-660642.
    ↩︎
  28. Aspinwall, Nick. 2021. “Taiwan Under Fire for Racist Policies Toward Southeast Asian Workers.” The Diplomat. https://thediplomat.com/2021/06/taiwan-under-fire-for-racist-policies-toward-southeast-asian-workers/.
    ↩︎
  29. Based on an interview with a former AEI colleague with prior work experience at the PRC-state-affiliated ASEAN-China Centre (中国—东盟中心).
    ↩︎
  30. “The Population Geography of Taiwan.” n.d. Department of Geography, National Taiwan Normal University. Accessed November 26, 2023. http://twgeog.ntnugeog.org/en/population_geog/.
    ↩︎
  31. Huang, SZ. 2018. “A Relationship Between Economic Performance And Air Pollution of a Science Park: A Case Study Of Hsinchu Science Park in Taiwan.” Applied Ecology and Environmental Research 17, no. 1 (October): 575-585. https://www.aloki.hu/pdf/1701_575585.pdf.
    ↩︎
  32. Chen, Marlene. 1998. “Does High Technology Mean High Pollution? – Pollution Control in the Semiconductor Industry.” Taiwan Panorama. https://www.taiwan-panorama.com/en/Articles/Details?Guid=18b82da5-420a-45cf-8f87-4cedf3d44fda&CatId=9&postname=Does%20High%20Technology%20Mean%20High%20Pollution%3F%20–Pollution%20Control%20in%20the%20Semiconductor%20Industry.
    ↩︎
  33. “Stable Supply of Natural Gas.” 2022. Energy Administration, Ministry of Economic Affairs, R.O.C. https://www.moeaea.gov.tw/ECW/english/content/Content.aspx?menu_id=8677.
    ↩︎
  34. Webster, Joseph. 2023. “Does Taiwan’s massive reliance on energy imports put its security at risk?” Atlantic Council. https://www.atlanticcouncil.org/blogs/new-atlanticist/does-taiwans-massive-reliance-on-energy-imports-put-its-security-at-risk/.
    ↩︎
  35. Li, Lauly, and Cheng Ting. 2022. “Taiwan’s frequent blackouts expose vulnerability of tech economy.” Nikkei Asia. https://asia.nikkei.com/Business/Technology/Taiwan-s-frequent-blackouts-expose-vulnerability-of-tech-economy.
    ↩︎
  36. James, Allisa, and John Loeffler. 2022. “Could a huge 90-minute power outage in Taiwan draw out the global chip shortage?” TechRadar. https://www.techradar.com/news/could-a-huge-90-minute-power-outage-in-taiwan-draw-out-the-global-chip-shortage.
    ↩︎
  37. Lin, TC, and Ian Huang. 2022. “How Taiwan’s expanding semiconductor industry deals with water shortages.” CommonWealth Magazine. https://english.cw.com.tw/article/article.action?id=3236.
    ↩︎
  38. Kelter, Frederik. 2022. “Will Semiconductors and TSMC Save Taiwan or Endanger It?” Foreign Policy. https://foreignpolicy.com/2022/11/09/tsmc-taiwan-battle-semiconductors-water-resource-scarcity/.
    ↩︎
  39. Lin, Christina. 2023. “Food resiliency—a soft spot in Taiwan’s silicon shield?” CommonWealth Magazine. https://english.cw.com.tw/article/article.action?id=3534.
    ↩︎
  40. Chiu, Hua-mei. 2014. “The Movement against Science Park Expansion and Electronics Hazards in Taiwan.” The Rise of Environmentalism, 15-22. 10.4000/chinaperspectives.6521.
    ↩︎
  41. Strong, Matthew. 2021. “Semiconductor project in north Taiwan held up by environmental concerns.” Taiwan News. https://www.taiwannews.com.tw/en/news/4137082.
    ↩︎
  42. Lin, Christina. 2023. “Food resiliency—a soft spot in Taiwan’s silicon shield?” CommonWealth Magazine. https://english.cw.com.tw/article/article.action?id=3534.
    ↩︎
  43. Kumar, Rakesh. 2023. “U.S. chip efforts have focused on advanced semiconductors–but low-tech legacy chips could give China an unexpected edge.” Fortune. https://fortune.com/2023/08/18/u-s-chip-efforts-have-focused-on-advanced-semiconductors-but-low-tech-legacy-chips-could-give-china-an-unexpected-edge/.
    ↩︎
  44. Arnold, Zachary. 2021. “Chinese Students Are Not a Fifth Column.” Center for Security and Emerging Technology. https://cset.georgetown.edu/article/chinese-students-are-not-a-fifth-column/. ↩︎
  45. Based on an interview with Tzi-Dar Chiueh, Dean of the National Taiwan University Graduate School of Advanced Technology.
    ↩︎
  46. Based on an interview with a White House National Economic Council staffer.
    ↩︎
  47. Based on an interview with Tzi-Dar Chiueh, Dean of the National Taiwan University Graduate School of Advanced Technology.
    ↩︎
  48. Miller, Chris. 2022. Chip War: The Fight for the World’s Most Critical Technology. New York: Scribner.
    ↩︎
  49. Based on interviews with senior officials in Taiwan’s National Security Council, Ministry of Economic Affairs, and Ministry of Foreign Affairs.
    ↩︎
  50. Taiwan National Statistics. 2023. “Taiwan: average monthly wage 2022.” Statista. https://www.statista.com/statistics/319845/taiwan-average-monthly-wage/.
    ↩︎

Author

Ethan Chiu

Ethan Chiu is a sophomore at Yale University studying Global Affairs and History. He currently serves as the 2023-2024 YRIS International Liaison. He has previously worked at the American Enterprise Institute, Department of Defense, and American Red Cross, and is currently a research assistant at the DOD Information Strategy Research Center and National Defense University.

ethan.chiu@yale.edu