Taiwan crisis: More important than oil: Taiwan’s semiconductors

Crises are like tutoring: After the 2008 banking crisis taught us how the financial markets work, the corona pandemic became a crash course in value chains. At that time, with semiconductor chips, not only were graphics cards and computers becoming scarce, but cars and refrigerators were also suddenly faced with long delivery times.

Two lessons stuck: There are now virtually no commodities that can be built without chips, and the global market for semiconductors is dependent on a single supplier – the Taiwanese Semiconductor Manufacturing Company. Although TSMC did not have to go into lockdown during the pandemic, since Western companies canceled their orders, the Taiwanese manufacturer had made new delivery commitments.

It is obvious that this dependence on semiconductors also has a geopolitical dimension. Or as historian Chris Miller, who teaches in Massachusetts, puts it in the introduction to his book “The Chip War”: “The USA still controls the market for silicon chips, which gave Silicon Valley its name. However, this supremacy is under threat. China now spends more every year on importing chips than on oil (…) To free the semiconductor industry from America’s stranglehold, the country is investing billions of dollars in developing its own chip technology (…) If this strategy is successful, Beijing will “We can transform the global economy and readjust the military balance.”

But before the historian Miller turns to his actual topic, namely this struggle for economic-geopolitical dominance, he first traces the history of semiconductor technology. He describes in a knowledgeable and detailed manner how, starting in the 1940s, transistors first revolutionized the arms race and space travel, then also the consumer goods industry, and how the spread of this technology was conditioned by competition between the USA and the Soviet Union, and later also between the USA and Japan.

In this first part, Chris Miller gets a little lost at times in the biographies of developers and entrepreneurs, whom he too enthusiastically idealizes as ingenious doer types. The book becomes really exciting when Miller approaches the present and thus the question of why global semiconductor production is actually concentrated in Taiwan today.

In this context, it is important that the USA is still the world market leader in PC processors via Intel. The big problem today are the chips that are used, for example, in military equipment, household appliances or cars. Depending on the application, specific semiconductor architectures are developed by engineering companies all over the world. As these chips become more and more compact – the semiconductor coating is sometimes only a few atoms thick – production is also becoming more and more demanding.

This is where Taiwan finally comes into play: From the beginning, TSMC’s business model has been limited to manufacturing chips designed by other manufacturers. For this reason, the Taiwanese company was not perceived as a competitor by other companies and was readily chosen as a contractor. More than half of global semiconductor contract manufacturing is now carried out in TSMC factories – for the most modern chips, the proportion is said to be over 90 percent.

This monopoly position is a huge problem for the global economy: a Chinese naval blockade, a severe earthquake in the region or a political crisis in Taiwan would bring industry worldwide to a standstill. But setting up comparable semiconductor production elsewhere in the world is, as Miller clearly outlines, far more complex than generally assumed.

Modern semiconductors are now produced using a photolithography process that uses extreme ultraviolet radiation (EUV). Miller describes the technical challenges with relish: “The best way to generate EUV radiation is to shoot a tiny drop of tin with a diameter of 30 micrometers through a vacuum at a speed of around 320 kilometers per hour. The tin is then irradiated twice with a laser, with the first pulse heating the tin and the second turning it into a plasma with a temperature of about a million degrees – many times the surface temperature of the Sun. This process of irradiating tin is then repeated 50,000 times per second to produce EUV radiation in the amount required to make chips.”

The exposure process developed by a US company and used by the Dutch company ASML requires high-precision lasers and mirrors, which in turn can only be built by other highly specialized companies in Germany, among others. The special achievement of TSMC is that it has assembled a large number of individual components into a functioning production system.

And that ultimately explains why semiconductor factories cannot be easily replicated. What is needed here is not only billions in investments, but also close business relationships with global partners. In the USA and Germany, attempts have been made to solve the problem by encouraging TSMC to set up plants in their own countries with the help of billions in subsidies. But as Miller notes, “None of the new factories will be dedicated to producing state-of-the-art chips, so TSMC’s most advanced technology will remain in Taiwan.”

The situation is even more complicated for China, which spends around 200 billion euros annually on imported semiconductors – more than on oil. President Xi has declared the development of his own semiconductor industry to be a strategic goal. But growing tensions with the West are standing in the way of too close cooperation between global market leaders and Chinese semiconductor producers. So Beijing not only has to build chip factories, but also highly specialized mechanical engineering companies that can, for example, produce the lasers required for the lithography process. That should take at least a decade.

Both the Chinese pathos about national reunification with Taiwan and the West’s democracy rhetoric obscure what is probably mainly about: If Beijing got the Taiwanese semiconductor industry under control, it would have enormous economic power. For this reason, the US must use all means possible to prevent the unification of China and Taiwan. At the same time, a war would also be devastating for everyone involved, because without the semiconductors from Formosa, the global economy would come to a standstill, which would likely lead to global social unrest. And finally, for Taiwan, the chip industry is, as President Tsai Ing-wen put it, “a protective shield made of silicon.” As long as the West needs Taiwanese semiconductors, the country will not be abandoned.

Chris Miller’s great achievement is that he explains these complex technological, business and geopolitical questions in a way that is easy to understand. Today’s world is characterized by both extreme interdependence and nation-state competition – a very dangerous mix. More than oil and raw materials, tiny technological goods are at the center of the dispute.

Chris Miller: The Chip War. How the USA and China are fighting for technological supremacy in the world. A.d. American English v. Hans-Peter Remmler and Doro Siebecke. Rowohlt, 500 p., hardcover, €30.