For centuries, wars have been fought to gain control over vital resources like gunpowder, saltpetre, coal and oil. Possessing these kinds of strategic commodities gives competing states industrial and military advantages over their rivals. The coming conflict between the US and China won’t be decided by oil, gas, finance or even nuclear weapons. Instead, the winner will be the country that controls one essential resource: advanced microchips.
That is the compelling premise of a new book by US professor of history Chris Miller. The product of extensive research and first-hand interviews, Chip War: The Fight for the World’s Most Critical Technology dives deep into a commodity used to produce many of capitalism’s most valuable and valued products. Its most important use, from the perspective of the major states and corporations, is in the production of advanced military equipment.
The book starts by tracing the history of microchips, right back to the early years of the Cold War. Computer processing began as a solution to the growing complexity of data management, mathematics and state bureaucracies. The invention of vacuum tubes made relatively instantaneous complex calculations possible. The problem with vacuum tubes, however, is that they were huge, unstable and prone to being attacked by moths (hence “debugging”).
Eventually, some cutting-edge scientists developed a way to replace these tubes with a switch called a transistor. These were placed on a base of either silicon or some other material, on which a circuit was etched. They were far smaller than the old tubes, and much less subject to mechanical or bug-related failures. Thus, the microchip was born.
Early chip companies struggled to find a market that could afford to pay for these expensive new devices. The US military quickly identified the potential of this new technology and became a vital source of funding and technical support for the fledgling industry. At the time, the US was locked in competition with the USSR, which achieved early victories, sending the first satellite to space in 1957, followed by astronaut Yuri Gagarin in 1961. Soviet exploits shocked the US establishment. It responded by directing enormous funds to research institutes and creating NASA in 1958. This had nothing to do with exploring the wonders of space; instead it reflected the consensus that rockets were the future of war.
The first chips were overwhelmingly directed towards the military-industrial complex. NASA’s Apollo missions were initially the biggest purchaser of chips, seen as necessary to get a man on the moon. Despite being primitive by today’s standards, these new chips were far more powerful and efficient than anything that had been seen before.
Around the same time, the US was seeking to update its first cruise missile, which had used reels of hole-punched tapes and a heavy onboard mechanical system for guidance. The chip-based system would be twice as powerful and half as heavy. Within a year, this one missile program alone absorbed 60 percent of all chip sales in the US.
Miller makes a compelling case that modern warfare is inseparable from the advances made in chip technology. The impact of US chip-making prowess was demonstrated in the First Gulf War, in which the most modern Soviet weapons were powerless in the face of America’s guided missiles and bombs. He cites a New York Times headline that captures the euphoria: “War Hero Status Possible for the Computer Chip”.
While this origin story makes for fascinating reading, it is served with a huge side of capitalist propaganda. Chris Miller is an ardent champion of US capitalism. Every chapter is infused with this right-wing perspective, to the point of blinding the author to the conclusions that emerge from his own work. For instance, Miller repeatedly insists that the US chip industry is an example of the superiority of America’s supposed free market system. This is despite the industry’s dependence on NASA and the US military for its very existence. In later chapters, he attacks the attempts by South Korea, Japan and China to replicate America’s success as state-centric, inefficient and somehow unfair. The USSR’s approach, which relied on the duplication of previous generations of US chips, is dismissed in even harsher terms and explained away as a product of the inherent limits of “communism”.
This follows a long tradition on the US right, which praises the virtues of the free market even as it aggressively deploys the state to support private capital (via protectionism) and the US empire (via wars). In recent years, the right has held up the development of the mRNA covid vaccine as a triumph of the American free market when it was underwritten by billions of dollars of US public funding and relied on research by the public National Institute of Health, which is legally barred from making profits.
Miller also praises the ruthless drive for efficiency that has characterised the chip industry. He celebrates the anti-union measures implemented by US companies in their domestic fabrication processes and the zest with which they sought out cheap labour via globalised production chains. Never mind that these were deliberately established in countries run by dictatorships that banned unions.
Less objectionable is the author’s appreciation for the feats of science and engineering involved in establishing and refining the chip production process. One memorable section describes the technology involved in the modern Extreme Ultraviolet (EUV) printing process, which requires a specially designed laser to pulverise 50,000 microscopic balls of tin per second. These explosions emit light that is captured by a complicated system of unimaginably perfect mirrors, which direct the light to imprint circuits on a silicon wafer. This incomprehensible technology allows for the etching of 114 billion transistors on the M1 chip designed by Apple for its latest smartphones. Chips produced in 1965 had just 64.
The technological advancement required to get to this point had significant economic implications. Marxists long ago identified that capitalist competition tends towards the concentration and centralisation of capital in larger and larger conglomerates, which can take advantage of economies of scale and pricing power to eliminate their rivals. While pure monopolies are rare, the domination of industries by a handful of companies is the norm. Nowhere has this been seen more than in the chip industry.
The Dutch company ASML is the only company worldwide that can produce EUV lithography machines. The most advanced microchips can be made by only two companies, Samsung and Taiwan-based TSMC. This extraordinary situation, where the most vital instruments of modern society are so tightly controlled, can be explained by simple economics. It cost more than $14 billion to research and commercialise EUV technology. The lithography machines themselves are hugely expensive, costing around $200 million each. Building a chip factory—called a fab—costs about $20 billion and requires highly specialised staff to operate. Very few companies can afford to invest these sums in the speculative hope of gaining market share. Each new round of research costs more to initiate, forcing more companies out of the market. Today, most major companies rely on TSMC to produce their high-end products, including early powerhouses such as IBM and AMD. Some, such as NVIDIA and Apple, never made chips themselves, focusing instead on creating effective new designs for chips that are then subcontracted to TSMC. Concentration is intense here too, with just three US software companies responsible for 75 percent of the designs generated. So much for the free market.
Miller belongs to the realist school of international relations, which (rightly) asserts that world politics is a zero-sum competition between rivals. He is an open supporter of US imperialism. Everything is seen through this lens. For instance, his chapters on Japan’s partially successful attempts at creating a chip industry are tinged with hostility—a legacy of the Reagan-era fear that Japan would supplant the US as the world’s economic powerhouse.
But China is subjected to the most sustained critique. It was interesting to learn that China was an early leader in the semiconductor field until its research programs were destroyed in the 1960s by Mao’s Cultural Revolution. Later, China adopted the USSR’s approach of stealing and copying US designs, which doomed it to lagging generations behind the cutting edge.
China has only recently developed a proper chip industry after state planners identified the problem of relying on the US and its allies for microchips. It now lavishes funding on the sector, from the research and development phase through to subsidising the establishment of fabs, usually at a cost far higher than commercial rates. In doing so, it has attempted to recruit industry experts to work for Chinese businesses and lure offshore companies to establish local fabs with the promise of tax breaks and other benefits.
Miller denounces China’s supposed systematic theft of intellectual property and attempts to “bully” corporations into building factories in China. (The US, of course, would never use its economic or military power to bully anyone.) His real concern is that some Chinese companies were using US tools to design and deploy advanced chips beyond what most US companies could. The most significant example of this was Huawei, which turned itself into one of the world’s largest and most profitable suppliers of 5G infrastructure.
In contrast to the hostility with which Miller treats the Chinese industry, he lavishes praise on Taiwan. Its success was based on an early decision not to worry about chip design or computer software but to focus entirely on fabrication, allowing Taiwanese companies to offer their services to every company on the planet without risk of intellectual property theft. This unique approach was brilliantly successful but relied entirely on state subsidies and contracts to get off the ground. It was made possible only by recruiting US scientists and companies to Taiwan. A cynical reader might think this double standard has something to do with Taiwan being a key US ally in the region.
The possibility of a future war over Taiwan has been discussed widely. But the book makes clear that producing advanced chips is a field that requires delicate machinery and a specialised workforce. There is no circumstance where the US would allow the highly strategic fabs—or the workers who run them—to be captured by China. In fact, as Miller explains in interviews about his book, Taiwan’s development of a chip industry was deliberately designed to make that country an irreplaceable part of US supply chains and security interests. Any interruption of TSMC’s production would devastate the world economy, which relies on Taiwan for over one-third of the computing power produced each year. This doesn’t preclude a war, for capitalism is a deeply irrational system. But it makes starting one a historic gamble.
Miller’s imperialist doublespeak makes sense only if we consider the context of the book’s publication. In the final chapters, he effectively describes the strategic threat posed by a rising China and the concentration of chip production in nearby Taiwan. He also lambastes the Obama administration for its failure to challenge China’s ascendancy. It is unsurprising, therefore, that Miller strongly supports the aggressive restrictions that the Trump administration placed on the sale of chips and chip-making equipment to China. These policies turned Huawei from one of the world’s largest sellers of technology to a business “fighting for its survival”, according to a recent memo by CEO Ren Zhengfei.
The Biden administration has taken Trump’s protectionism to a new level. Biden has expanded sanctions substantially, banning not only the sale of chips made by American companies but any chips made by companies that used technology produced by an American company. These sanctions are no longer justified by the allegedly bad behaviour of specific companies but by the very nature of China as a geopolitical threat to US supremacy.
The recent decision to ban ASML from selling its advanced chip-manufacturing machines to China is the most devastating yet. Not only does it foreclose the possibility of China producing cutting-edge chips using EUV processes, but it also bans the most advanced technology from a generation earlier. This condemns China to a future in which it lags years, if not decades, behind the US. ASML protested initially (the ban will cost it billions) but has little option but to comply, given that the US controls its San Diego subsidiary—the sole intellectual property owner of the EUV process.
All of this is a reminder that despite the chatter about US decline, it remains a uniquely powerful actor on the world stage. Though its manufacturing sector has declined to some extent, in research and development it is largely unrivalled. Its dominance of global finance means it can impose crippling sanctions unilaterally.
These punitive measures against the Chinese tech industry should not be understood as just another escalation in simmering trade tensions. They are a declaration of economic war. But not everyone in the US is happy about this. Apple and NVIDIA have decided to make chips with the explicit goal of bypassing export bans, and plenty of tech companies are still investing in China. Miller points out that while the US military was central to establishing the chip industry, it is now a bit player in the global market for chips, which has long been dominated by consumer goods. China, of course, is the world’s largest consumer goods market.
Miller is smart enough to identify the implications this has for any decoupling between the US and China. “The entire chip industry depended on sales to China”, he writes, before quoting a US chip executive: “Our fundamental problem is that our number one customer is our number one competitor”. We should not place too much faith in this interdependence as an antidote to war: Germany and Russia were among each other’s largest trading partners prior to each of the world wars.
Another challenge to the sanctions and decoupling narrative is that it is impossible to track the location of every chip: they can be sold for one purpose and then used for another. For instance, a recent report found that advanced NVIDIA graphics cards that had long been banned from China because of their potential use in nuclear research have been found in Chinese nuclear research labs.
Chip War is not an objective study of the microchip; it is a dangerous intervention into US politics in favour of a policy of total economic and political war against China. In researching the book, Miller was granted access to senior levels within the security establishment in Washington. Precisely because it seeks to make a persuasive and rigorous case on behalf of these figures, there is much to be learned from its systematic account of some of the world’s most important geopolitical and industrial fault lines. Anyone seeking to understand the present and future shape of imperial conflict needs to read and absorb its arguments.
