By Finn Watson
For governments around the world, securing domestic production of high-grade microchips is a vital security interest that has prompted subsidies, trade restrictions, and fierce international competition.
The Technological Race
The economic ascendency of microchips is truly remarkable. Since the invention of the transistor in 1947, which facilitated the invention of the integrated circuit, or microchip, in 1958, they have permeated nearly every object and industry. From the device you are using to read this article, to every new car you see, to the most advanced military applications such as jets and missiles, microchips have enabled a massive expansion in computing power and facilitated the technological boom of the past half century. Gordon Moore, co-founder and former CEO of Intel, observed that the number of transistors in an integrated circuit doubles about every two years. In a trend that has largely been upheld since his observation in 1965, ‘Moore’s law’ has facilitated the doubling of computer power and subsequent decrease in electricity consumption that has enabled the computerization of the economy and the rise of technologies that have added immense value to the economy.
However, the exponential growth in transistor count presents a serious logistical and technological challenge: the more transistors one attempts to fit into a defined space, the smaller these transistors must be. Furthermore, for the chips (which constrain billions of transistors) to become faster, they must get smaller. Thus, over fifty years of Moore’s law, and as of 2022, transistors are remarkably small. In the second half of this year, Taiwan Semiconductor Manufacturing Company, or TSMC, is beginning production of the 3nm process which has an astonishing transistor density of 250-300 million transistors per square nanometre. With experimental transistors reaching one-third of a nanometre in length – as thick as a single layer of carbon atoms – there are concerns that engineers are reaching the ‘hard limit’ of how much smaller these microchips can become. For decades now, pundits have speculated on the “death of Moore’s law” and the inevitable impossibility of further breakthroughs in computing power. Despite these doubts, businesses and governments are going full speed ahead in pursuit of better, faster, smaller, chips.
As one might suspect, these microchips require immense and sophisticated industrial facilities as well as significant capital investment. For example, the semiconductor fabrication facilities or ‘fabs’ need to be seismically insulated from minute vibrations in the earth’s crust which would interfere with the delicate machinery involved. The production must also be done in clean rooms to prevent hair and other contaminants from rendering the silicon wafers unusable. Furthermore, an important step in the microchip production are lithography machines that are almost exclusively produced by one company, ASML, and cost from $150 million to $400 million each. All considered, TSMC’s new fabrication facility which is going to be the sole production facility of the aforementioned 3nm chip is expected to cost around $20 billion dollars. It is for these engineering challenges and massive required capital expenditure why the barrier of entry for companies looking to create microchips is so high. In fact, as of 2021, only TSMC and Samsung produce the advanced chips in the 10nm-5nm category, and of these over 80% are produced by TSMC. In this way, TSMC produces nearly all of the high grade chips making it a lynchpin in the global economy and an essential security concern for many.
The ascendency of TSMC is further punctuated by the transition of established US hardware titans such as Intel and Qualcomm to the ‘fabless’ manufacturing model wherein the sale and design of chips are done in-house, and manufacturing is outsourced to ‘foundries’ such as TSMC. Especially with the newest generation of microprocessor manufacturing, with its astronomically high capital investment figures, US producers have fallen behind. In 2018, the US’s largest domestic foundry, GlobalFoundries exited from the production race in the new chipsets. Furthermore, Intel, once a titan of manufacturing innovation, fell under significant investor criticism after losing the edge to TSMC in terms of manufacturing ability. This move away from vertically integrated micro processing businesses and towards an industry where production is dominated by TSMC represents an emerging vulnerability for governments and is attracting significant political attention.
The Geostrategic Race
A United States Commission on National Security and Artificial Intelligence (NSCAI) report highlighted the vulnerability of America vis-a-vis its reliance on Taiwanese microchip manufacturing. The report stated:
“The United States lacks domestic facilities capable of producing, integrating, assembling, and testing SOTA microelectronics at scale… which are critical for national security applications. This lack of access forces the USG, as well as the U.S. defense industrial base and other commercial U.S. firms, to rely on foreign fabrication and complex global supply chains for production, which increases risks to product integrity and exposes U.S.-developed intellectual property (IP) to theft. According to a DoD official, this gap should provoke a “graceful and considered kind of panic” among U.S. policymakers.”
Considering the speed and scale of US recent congressional support, it is evident that a ‘graceful and considered kind of panic’ has certainly descended on Washington’s strategists. With uncharacteristic speed and efficiency, congress has enacted sweeping protectionist export controls, subsidies, and other competitive measures to address this strategic concern. The largest of which is the $280 billion dollar “CHIPS and Science Act” that provides $52 billion towards boosting domestic production of microchips and energizing R&D initiatives. The act also provides several billion in support of legacy chip production and hefty tax credits for advanced manufacturers. Furthermore, the US has implemented export controls that prevent China from accessing high-grade chips, the tools to make them, and the US talent that designs them. In addition, the US has invoked the foreign direct product rule (FDPR) which prevents companies, regardless of where they are based, from selling advanced chips to Chinese affiliated entities if the companies use American technology in the production process. As virtually all advanced chips use technologies derived from US research, the use of the FDPR has sweeping negative implications for Chinese firms and strategists. These extensive policies are designed to protect the US’ supply of microchips by bolstering domestic production as well as maintaining their competitive edge by regulating Chinese access to chips. In short, US policy is to slow down Chinese access while subsidizing domestic production. In this way, chips are coming to define Sino-American competition.
Recently, contests between the US and China over Taiwan have heated up. As highlighted by House Speaker Nancy Pelosi’s visit to Taiwan in which the Chinese government threatened a possible intervention by the PLA Air Force, defending the independence of Taiwan is a vital goal of US policymakers. America’s reliance on the island’s microchip production is undoubtedly a core motivation for the visit as demonstrated by Pelosi’s tour including a meeting with TSMC’s executive chairman. Moreover, the de facto independence of Taiwan is so strategically important that the Biden Administration has moved away from an approach of ‘strategic ambiguity’ to outright stating that the US military would defend Taiwan if China were to invade. In this way, the ascendency of TSMC and Taiwan in microchip production is fueling Sino-American antagonisms over the island nation and motivating Washington’s strategic shift to the Indo-Pacific.
However, sweeping financial and legal subsidies are not limited to the west. In character, China has engaged in the most aggressive investment campaigns in support of its domestic microchip industry. For example, in firms across the US, Korea, and Taiwan, state support as a percentage of revenue is less than 5%, but in China’s largest semiconductor firms such as Tsinghua and SMIC, subsidies can be upwards as high as 40%. Furthermore, an OECD report on market distortions in international markets describes how no countries except for China offer below-market financing to enterprises. Despite this intense government support, China remains a generation behind US-Taiwanese chips. However, whether this investment effort will result in China overtaking the US and Taiwan or if it’s another example of Chinese malinvestment is yet to be seen.
The geostrategic race for microchip goes beyond just China and the US as domestic security and technological advancement in chips is a key strategic concern of the European powers as well. Similarly to the US, much of Europe’s production capacity has been offshored to Asia and its firms have embraced the fabless production model. However, whereas the US still maintains around 30% of the world’s production capacity, Europe accounts for around 10%. On top of this, the chips that Europe is producing are of much lower specifications. European firms account for only 1.7% of production in the 10nm-20nm category. For these reasons, Europe has identified this reliance on foreign production as a point of strategic interest and moved to pass the European Chips Act. The European Chips Act, much like its American counterpart, earmarked €43 billion in public and private investment to bolster regional manufacturing abilities. However, unlike the US, Europe has not yet incited any of the large Taiwanese manufacturers to open fabs in the EU. Although the European Chips Act is a step in the right direction, Europe remains more vulnerable than the US. As one might expect, the EU and UK have aligned their policies with the US and against China. When the Dutch company Nexperia moved to acquire Britain’s largest fab, the Department for Business, Energy and Industrial Strategy issued an order that forced the company to sell 86% of their stake due to national security concerns. This was because Nexperia, although based in the Netherlands, is owned by Wingtech, a Shanghai company backed by the CCP government. The government cited the ‘cluster of expertise’ in Wales that would be vulnerable to subversion due to its links with defence contracting. Such a move has come under criticism for threatening to erode the already soured Sino-British relationship. Thus, for governments around the world, the microchip industry has become a key arena for competition that will come to define the balance of power.
Whether we have realized it or not, microchips have become a ubiquitous and essential resource. For governments around the world, securing this resource is a defining aspect of their foreign policy. As far as chip manufacturing is concerned, Taiwan is the center of the world and as long as governments are drawn in by its’ gravity, clashes are inevitable. Who will win the race for microchips has yet to be seen but the participants are certainly going to give it their all.
The views expressed in this article are the author’s own, and may not reflect the opinions of The St. Andrews Economist.
Photo by Laura Ockel via unsplash.com