PDF Summary:Chip War, by Chris Miller

In Chip War, economic historian Chris Miller explores the rise of the semiconductor industry. Through his analysis and recounting of this history, Miller shows how semiconductor chips became the delivery mechanism for the binary code that powers our digital world—from personal computers to household appliances to advanced weapons systems. As Miller writes, chips have become a linchpin of geopolitical power and a resource that powerful actors, from entrepreneurs to national security strategists and politicians, have sought to control.

In this guide, we’ll explore the founding of the chip industry in the US, the emergence of Japan and Taiwan as major centers of chip design and manufacturing, and the rise of China as a new geopolitical rival in the global semiconductor competition—sparking concerns among Western firms and governments as national security interests became intertwined in the chip industry.

Throughout the guide, we’ll supplement Miller’s analysis with insights from other experts and analysts.

(continued)...

American support: The United States provided financial aid and resources to Japan through programs like the Marshall Plan, which helped Japan rebuild its infrastructure and industries. In addition, the US facilitated trade relations with Japan, providing access to American markets for Japanese goods. This access allowed Japanese industries to export products and generate foreign exchange, which was crucial for economic recovery.

The Japanese Advantage

By the 1980s, writes Miller, Silicon Valley faced a formidable challenge from Japan. Japanese companies had not only caught up with American technology but were also no longer just licensing American technology—they were producing their own chips, and higher-quality ones at that. Meanwhile, innovations like the Sony Walkman, first introduced in 1979, underscored Japan's prowess in consumer electronics.

Miller notes that Japanese chip companies also benefited from a confluence of macroeconomic forces that worked in their favor at this time. Thanks to low inflation and a booming economy, Japan had a glut of savings in the late 1970s, which was used to fund investments in new industries like semiconductor chips and consumer electronics. Because there was so much capital, firms could access business loans at low interest rates, fueling quick growth. At the same time, the Japanese government implemented protectionist policies, including tariffs and quotas on foreign chip shipments. This further tilted the playing field in Japan's favor.

Japanese Industrial Policy and the Semiconductor Industry

Some researchers assert that the Japanese government’s support for its booming semiconductor industry at this time extended well beyond the tariffs and quotas that Miller describes. Japanese industrial policy in the semiconductor industry played a pivotal role in the global market share gain of Japanese semiconductor companies. Notably, there was a significant increase in funding for research and development (R&D) in semiconductor manufacturing equipment, which grew to 26% of Japan's total R&D spending by 1977, up from just 2% at the beginning of the 1970s.

The Japanese government also invested $300 million in the establishment of the Super LSI Technology Research Association, a public-private technology research project involving Japan's six main computer companies. This collaboration allowed these rival companies to work together and share information, fostering innovation and the development of a common technology platform.

This close state-private coordination propelled Japanese firms to dominate the global semiconductor market, with Japanese companies accounting for 51% of worldwide sales by 1988.

The American Response

Miller writes that, in the face of this challenge, American semiconductor CEOs realized they needed both government support and new technologies to regain their former position.

Advocating that chips were a strategic resource vital to American military and economic security, semiconductor industry heads founded Semiconductor Manufacturing Technology (SEMATECH) in 1987. SEMATECH was a consortium of semiconductor manufacturers, equipment suppliers, and the US government. Its primary mission was to advance semiconductor manufacturing technology and enhance the competitiveness of the US semiconductor industry by facilitating coordination between the Pentagon and the industry to enhance production and collaboration.

(Shortform note: While SEMATECH did achieve success by fostering collaboration among its member companies, by the mid-2000s, the semiconductor landscape had evolved significantly. New challenges emerged, including the globalization of the semiconductor supply chain and the rise of semiconductor manufacturing in Asia. SEMATECH’s funding model also faced challenges. It relied on contributions from member companies and government support, but sustaining funding levels became increasingly difficult. Eventually, the consortium was absorbed by the State University of New York Polytechnic Institute in 2015.)

The Japanese Begin to Falter

By the 1990s, writes Miller, the Japanese position in the semiconductor industry began to decline.

In addition to the new American technology and improved public-private coordination described above, economic events in Japan contributed to the nation’s relative decline in the semiconductor space. During the 1990s, the Japanese economy took a downturn, while the United States emerged as the dominant economic force of the decade.

Miller attributes this decline in Japan's economic prowess to several factors, including an excess of cheap capital and overinvestment. With such easy access to capital, Japanese firms found themselves less compelled to compete on the basis of quality. It made more sense for many Japanese companies to churn out commodified and generic chips—which were already being produced more affordably and effectively by South Korean companies like Samsung and American firms like Micron. While profitable for a time, this complacency and loss of competitive edge left the Japanese chip industry in a weak position to adapt to coming technological changes.

Japan’s Lost Decade

The period of relative Japanese economic decline Miller describes is known to economic historians as “Japan’s Lost Decade”: a period of economic stagnation and financial crisis that extended from the early 1990s into the 2000s. The era was marked by a number of economic events and trends:

• Asset price bubble: The roots of Japan's economic troubles can be traced back to the late 1980s when the country experienced an asset price bubble, particularly in the real estate and stock markets. Speculation drove up property and equity prices to unsustainable levels—until the bubble burst, triggering a financial crisis.

• Credit crunch: The financial crisis led to a credit crunch, where banks became reluctant to lend, and businesses and consumers faced difficulties accessing credit.

• Deflation: One of the defining features of Japan's Lost Decade was deflation, a persistent decrease in the general price level. Falling prices discouraged consumer spending and business investment, as people expected goods and assets to become cheaper in the future.

Part 3: The Rise of TSMC and the Foundry Model

Up to now, we’ve explored the rise of the semiconductor industry in the United States and the robust challenge to the American industry from the Japanese that began in the 1960s. In this section, we’ll turn our attention to the rise of Taiwan Semiconductor Manufacturing Company (TSMC) and the Taiwanese chip industry beginning in the 1980s, the resulting shift to the foundry model for chip manufacturing, the corresponding shift to the “fabless” model for chip design, and the decentralization and globalization of the chip supply chain.

Morris Chang Founds TSMC

In 1985, Chinese-born entrepreneur Morris Chang left Texas Instruments (TI) after being passed over for the CEO role. Spurned by the company to which he’d devoted his career, Miller writes that Chang decided to seize a different opportunity—to make Taiwan the new epicenter of the chip industry and place himself at its head. In 1987, he founded TSMC, a public-private partnership that would become the dominant global player in chip manufacturing.

Fortunately for Chang’s ambitions, writes Miller, the Taiwanese government was eager to develop a domestic semiconductor industry to reduce its reliance on foreign technology and create a strategic economic advantage. Thus, they were enthusiastic about providing Chang’s initiative with significant support through financial incentives, access to resources, regulatory assistance, and tax incentives. Indeed, this backing was crucial for establishing the necessary infrastructure and acquiring advanced manufacturing equipment for large-scale chip production.

TSMC Aims to Maintain Dominance by Globalizing Production

As its domestic semiconductor industry has grown, TSMC has maintained a strong desire to safeguard its competitive edge. One way of doing this is by investing in chip facilities in the US, which would help the company diversify its production locations and enhance supply chain resilience.

This strategy has been a flashpoint of controversy in Taiwan. In 2022, the state-backed firm announced plans to invest $40 billion in a new advanced chip manufacturing facility in Arizona. However, Taiwan’s economy and national security are closely tied to the semiconductor sector, and TSMC's investments in the US have sparked concerns in Taiwan about technology and intellectual property leakage to the US and possible future dependency on US facilities.

The Foundry Model

Miller observes that what truly set TSMC apart from other chip companies was its pioneering role in spearheading what’s known as the foundry model.

Instead of designing their own chips, TSMC only manufactured chips designed by other companies—they were a manufacturing workshop. This offered distinct business advantages for TSMC through economies of scale and the opportunity to hone its production capabilities.

The Advantage of Economies of Scale

Miller writes that by mass producing chips for multiple customers, TSMC could achieve cost efficiencies through economies of scale. A small chip manufacturer faces relatively high costs to manufacture each chip—they have to set up and run their manufacturing equipment, pay for labor, and cover other fixed costs, with those costs spread out over a small number of chips. Thus, if the small manufacturer spends $100,000 on those startup costs but produces only 1,000 chips, their cost-per-chip is $1,000. But a large-scale mass producer like TSMC might produce 10 million chips, dropping their fixed costs down to $0.01 per chip.

The Limits of Economies of Scale

Although economies of scale do offer significant advantages for large enterprises, there are limitations. As companies increase their production levels, there is a point at which cost savings from economies of scale may no longer be achievable. Beyond a certain production threshold, cost reduction may not continue.

For example, a company that manufactures smartphones will benefit from increasing its production volume as it negotiates better deals with suppliers, optimizes its manufacturing processes, and reduces per-unit production costs. However, the company is eventually operating at maximum capacity, and any further increase in production would require significant new investments in facilities and equipment.

TSMC’s Production Edge

Moreover, by focusing relentlessly on production, TSMC was able to optimize its manufacturing processes, invest in new research and development, and purchase cutting-edge lithography equipment at scale by buying in bulk. All of this made its manufacturing processes more efficient—leading to faster production times and ever-lower labor costs per chip.

(Shortform note: Beyond bulk buying, TSMC’s scale affords it other advantages that contribute to its production edge. In particular, TSMC's ability to stockpile critical materials, such as silicon wafers, during periods of high demand or shortages, enhances its supply chain resilience. This helps ensure a steady production flow even when the company is faced with external supply disruptions.)

TSMC’s Dominance

By achieving scale like this, TSMC was able to produce chips at a price that its competitors couldn’t beat.

TSMC secured a hammerlock on the manufacturing process for key electronics like iPhones and other smartphones—achieving this position after Intel famously turned down Steve Jobs for the iPhone contract, not seeing the potential of the smartphone market. With major customers like Apple, TSMC held 50% of the global chip foundry market by 2015.

(Shortform note: In order to maintain its globally dominant position in the chip industry, TSMC has had to strategically expand production into other regions. However, TSMC is committed to preserving its Taiwanese identity. The company's origins and headquarters are in Taiwan, where it has a significant presence and continues to play a vital role in the country's economy. The company's expansion into other countries, such as the United States and Japan, is driven by the need to diversify its production and reduce geopolitical risks but not to sever its ties with Taiwan. The company's presence in Taiwan remains integral to its identity and the broader semiconductor landscape, despite its strategic expansion to other regions.)

The Fabless Model

Miller writes that TSMC’s success also heralded a revolution in chip design, as the chip design industry shifted toward what’s become known as the “fabless” model—in which the company focuses only on chip design rather than design and manufacturing.

TSMC and other chip manufacturers enabled chip designers to outsource their manufacturing operations to companies engaged solely in manufacturing. By thus freeing themselves from the high startup costs of chip manufacturing, these designers can achieve the same optimization and efficiency in chip design that TSMC achieved in chip manufacturing.

(Shortform note: Despite the advantages of the fabless model, fabless semiconductor startup companies do face limitations and operational challenges. By relying on third-party foundries for chip production, fabless companies can face increased production costs and capital constraints, making it challenging for them to compete with integrated semiconductor firms. In addition, depending on external foundries for chip manufacturing introduces supply chain risks, as any disruptions or capacity constraints at the foundries can impact the production and delivery of their chips. Finally, limited control over the manufacturing process can affect the quality and consistency of fabless companies’ chips.)

The Decentralization and Globalization of the Chip Supply Chain

In the 2000s and 2010s, more designers began to shift to the “fabless” model, in which these companies no longer needed the fabrication equipment to manufacture their chips. Miller writes that this shift was a major factor in the globalization and decentralization of the chip supply chain—with the United States holding a smaller share of chip manufacturing as that side of the industry moved to Taiwan and other East Asian countries.

(Shortform note: Recognizing the declining role of the US in chip manufacturing, American policymakers have sought to strengthen the country’s position in the global chip supply chain. Specifically, the US government has allocated significant funding to bolster domestic chip manufacturing. This includes the 2022 CHIPS Act, which provides $52 billion in incentives for semiconductor manufacturing and research. The aim is to reduce dependence on foreign chip production and reduce vulnerabilities in the supply chain, especially in times of global crises.)

Part 4: The Emergence of China

So far, we’ve explored the origins of the chip industry with a handful of American entrepreneurs after World War II, the emergence of the Japanese semiconductor industry in the 1960s, the rise of TSMC and the foundry and fabless models, and the globalization of chip manufacturing and design.

In this section, we’ll turn our attention to the increasingly prominent role of the People’s Republic of China in the global chip game. Specifically, we’ll look at China’s early attempts to establish an onshore chip industry; the recognition by Chinese leadership of the importance of semiconductors for the nation’s economic, military, and national security future; and the geopolitical concerns and challenges that China’s emergence as a major player in the chip industry has for the US and its allies.

China’s Semiconductor Ambitions

According to Miller, in the late 20th and early 21st centuries China embarked on a significant journey to emerge as a global player in the semiconductor industry.

The Chinese government had a vision of making mainland China a global semiconductor powerhouse. They believed China’s low labor and manufacturing costs could lure semiconductor investment to the country, play a key role in the nation’s recovery from the radicalism and chaos of the Maoist era, and be the stepping stone for China to finally play a central role on the global stage.

The Chinese Economic Reforms

To more fully understand China’s push to develop a domestic semiconductor industry, it’s worth exploring the historical context in which it’s taken place.

China initiated a series of economic reforms in the late 20th century, notably under the leadership of Deng Xiaoping. These reforms began in the late 1970s and continued through the 1980s. In 1980, China designated Shenzhen as its first Special Economic Zone. These zones were created to attract foreign investment by offering favorable policies and fewer restrictions to foreign companies. At the same time, China embarked on a path of trade liberalization, reducing trade barriers and making it more enticing for foreign companies to invest. The country also made substantial investments in education and workforce development throughout the late 20th century, aiming to cultivate a skilled labor force. These efforts continued through the 1990s and beyond.

As a result of these reforms and the influx of foreign investment, China experienced rapid economic growth, becoming one of the world's largest economies. This growth trajectory began in the late 20th century and continued into the 21st century.

China’s Early Challenges

However, observes Miller, China’s early ventures faced formidable challenges. Although Chinese cities like Zhengzhou and Dongguan emerged as hubs for chip and smartphone assembly, they were concentrated on the lower end of the value chain—providing largely unskilled labor to assemble iPhones and other devices for foreign companies like Taiwanese giants Foxconn and Wistron.

Despite the success of gaining a foothold in the low-value end of the chip value chain, China's semiconductor market share still trailed that of geopolitical rivals such as the US, South Korea, Japan, and, most significantly, Taiwan.

The Opportunities and Challenges of China’s Unskilled Labor Pool

Miller notes that the main resource China could contribute to the global economy at this time was its vast pool of unskilled labor—indeed, this once provided a significant advantage for China’s manufacturing industries, allowing China to become the "world's factory" by offering low-cost labor for various production processes. As China's economy evolved and technology advanced, there was a shift toward a more skilled and educated workforce. The country invested heavily in education and vocational training to develop a workforce capable of handling advanced manufacturing and technology-related jobs.

However, many industries that relied on low-skilled labor faced rising labor costs as workers demanded higher wages and better working conditions. The shift away from unskilled labor impacted industries like textiles, apparel, and simple assembly, which had traditionally thrived on low-wage workers. Some of these industries began relocating to countries with lower labor costs.

The Push for Geopolitical Dominance

Miller notes that the Chinese leadership—notably General Secretary of the Chinese Communist Party Xi Jinping—recognized that China's economic, military, and national security future depended on developing a homegrown, high-value semiconductor industry. They needed to move beyond assembling devices for global tech giants like Apple and Samsung and begin leading in chip design and manufacture. Beyond the economic benefits, there were also national security and geopolitical strategic concerns behind building a domestic chip industry—key technologies underpinning China's massive surveillance state heavily relied on inputs of foreign-made chips. If those supplies were suddenly cut off, China would be at risk.

In the 2010s, China began pouring tens of billions into its information technology sector, pushing both private investors and state-owned enterprises to invest and shift the center of technological innovation eastward.

(Shortform note: China's efforts have faced resistance from various quarters, including the United States and its allies. The United States has imposed sanctions on Chinese semiconductor companies and restricted their access to advanced technology, particularly American-made equipment and software. These sanctions aim to curb China's rapid advancement in semiconductor manufacturing. In addition, the US is working to strengthen alliances with like-minded countries to collectively counter China's semiconductor ambitions. This includes efforts to coordinate export controls and technology restrictions.)

China Targets Western Firms

Miller writes that, on the heels of China’s infusion of state investment into its domestic chip industry, Western firms, eager to get a foothold in the massive and lucrative emerging Chinese market, began doing more and more business with China. Companies like AMD, IBM, Qualcomm, SoftBank, and Arm began licensing core technology to Chinese companies or entering joint ventures with Chinese government-backed enterprises. While this arrangement has been profitable for these companies, critics raised the alarm that these companies’ eagerness to tap into China gave the Chinese authorities leverage over them. The concern is that the Chinese could exploit this leverage to gain access to important technology with national security implications for the US and its allies.

Meanwhile, Chinese interests also sought to buy stakes in US semiconductor companies, further complicating the landscape. Chinese mega-firms like Huawei rose to prominence, grabbing significant global market shares through replication and, at times, alleged technology theft, while also leveraging their position as major customers of massive suppliers like TSMC.

TikTok and the Global Concern About Chinese Espionage

Concern about Chinese technology intrusion extends beyond the semiconductor industry. The White House, Congress, US armed forces, and more than half of US states have banned the Chinese-owned social media app, TikTok, from government-issued devices. There’s broad concern that TikTok’s parent company, ByteDance, could share Americans’ personal data with the Chinese government, or push Chinese propaganda and misinformation.

TikTok has strenuously denied sharing user data with the authoritarian government, and in fact, studies from 2023 and 2021 found TikTok’s data collection practices to be comparable to—and no more of a threat than—those of US social media apps.

Many experts argue that a total TikTok ban is problematic. Civil liberties groups say a nationwide ban likely could create a dangerous precedent for the US government to dictate how Americans communicate—negatively impacting millions of influencers and other US citizens whose identity, self-expression, communications, and financial health are tied to the app.

The Global Chip Competition

Miller writes that these developments showcase the complex interplay of economics, technology, and geopolitics in the semiconductor industry. Computing power, the ability to produce chips and complex systems, and the capacity to transmit data faster and more accurately are the modern keys to geopolitical influence and military prowess.

The semiconductor industry had become a focal point in the US-China rivalry, and the stakes were higher than ever. The world was beginning to understand that the struggle for dominance in this industry was not merely an economic or industrial issue—but a matter of national security and global power.

Is the US Winning the Chip War?

Some experts argue that in the global semiconductor competition between the United States and China, the US is currently leading. Several key factors contribute to the US's advantage in this race:

Technological innovation: The US remains a hub for cutting-edge semiconductor design and development. Companies like Intel, NVIDIA, and Qualcomm continue to drive innovation in chip technology, ensuring the US maintains a significant lead in creating advanced semiconductor solutions.

Export controls: The US government has implemented stringent export controls that limit China's access to advanced semiconductor technology. These controls restrict the sale of chips, chip-making equipment, and software containing US tech to China, significantly hampering China's semiconductor progress.

Domestic investment: The US has initiated programs like the CHIPS Act, offering substantial grants and subsidies to companies involved in semiconductor manufacturing within the US. Meanwhile, major players like TSMC and Micron are investing billions in new facilities in the US, leveling the playing field against Asian competition.