‘Microchips are the new oil.’ That saying is now increasingly echoed among national security advisors, policymakers, and geopolitical experts. What are microchips? How did we get here? And why is it so crucial in determining the security and stability of our world?
To answer these questions, OPR Interviews Editor Jason Chau spoke to Chris Miller, Associate Professor of International History at the Fletcher School at Tufts University. He is the author of Chip War: The Fight for the World’s Most Critical Technology and winner of the Financial Times Business Book of the Year Award 2022. In our interview, we talks about the state of technological competition between the West and China, the development of microchips in the past decades, and the global reliance on Taiwan, a place situated at the very center of the most consequential geopolitical divide of our time.
As someone with a background in studying Russia and Eastern Europe, what drew you to the topic of semiconductors and this technological competition between China and the West?
I was interested in the evolution of military technology during the Cold War, and I was puzzled by the fact that the Soviet Union was capable of producing key defense technologies of the early Cold War, namely atomic weapons and long range delivery systems like ICBMs (intercontinental ballistic missiles), but was unable to produce some of the key types of defense systems with long range precision strike capabilities used at the end of the Cold War.
I tried to start by understanding that question: what went wrong in the Soviet defense industrial process? It quickly emerged that the key factor that had changed the types of systems that the US was acquiring over the course of the Cold War was the application of computing power to military systems. That was possible only because of semiconductors, which were the device that made it viable to miniaturize sufficiently powerful computing systems in many different types of military and civilian systems. And so I came to realize that you cannot understand the evolution of military power over the past 75 years and the shifts in technological or economic influence without paying attention to that.
Can you highlight, if there is a massive disruption of the supply in semiconductors – microchips – how severe will the consequences be to the world? What effects will that have on militaries and beyond? Would whole economies essentially stop running given how ubiquitous semiconductors are?
The main effect would be beyond, because the entire economy today runs on cement conductors. It’s not just smartphones and PCs, it’s coffee makers and dishwashers, microwaves, automobiles.
If we were to have a major disruption, I think the most plausible disruption of any scale would be some sort of military escalation between China and Taiwan. That’s because Taiwan produces 90% of the world’s most advanced processor chips, and over one third of the computing power the world adds each year. If we lose access to all those, the world would be plunged into a manufacturing crisis unseen since 1929, measured in the trillions of dollars. It would trigger a deep recession, with shortages of all sorts of goods, especially electronics.
Can you briefly talk about how [the] microchip was developed in the US in the 1950s, but its production was shifted to East Asia and specifically Taiwan?
The Taiwanese emergence as a major chip producer has its origins in the 1980s when the Taiwanese government negotiated a deal with Morris Chang, who would later found TSMC (Taiwan Semiconductor Manufacturing Company Limited) to help the country build up a chip industry. But it wasn’t really until the 1990s and 2000s that TSMC began to grow.
It grew for a number of reasons. One: because of the brilliant business sense of Morris Chang. Two: he had a unique business model whereby he wouldn’t design any chips, he would only manufacture them, which allowed him to serve a wide array of customers. Three: its impressive ability to integrate itself with the world’s semiconductor ecosystem. Although TSMC is the world’s most advanced chipmaker, it still has to buy equipment like chip-making and software tools from other countries to integrate itself with chip designers.
Do you think the Taiwanese government played a big part in making sure that the semiconductor industry flourishes in Taiwan? Do you think there were active industrial policies that made a difference in this case?
There’s no doubt the Taiwanese state played a large role here. It was the largest initial investor in TSMC – almost three quarters of the initial investment came either from the government or from Taiwan’s businessmen whose arms were twisted to invest by the government. I wouldn’t say the state is the reason TSMC succeeded, but the shape of TSMC today is largely attributed to the Taiwanese government’s early support.
Compared to companies from other countries or other companies, how far ahead is TSMC in its technological capabilities?
We can mention Samsung in particular, and to a certain extent, Intel. But the question to ask is not just about who can produce the most advanced chips, but also who can produce chips for which customers. When you design a chip, you design it with a certain set of specifications that are suited to a specific manufacturer. If you’re a company like Apple, which has contracted with TSMC to produce most of its chips, it wouldn’t be an easy thing to move your designs to another company.
TSMC not only has very advanced technology, it also has an extraordinary capacity to manufacture. It sets the tune by which the rest of the industry follows. It’s the sun around which the rest of the solar system is orbiting because it’s everyone’s most important supplier or customer or partner.
As you’ve mentioned in the book, China is spending a lot of money to boost their homegrown semiconductor industry. Because of all these investments, do you think China has surpassed or is on an equal playing field with the West, specifically the US on semiconductor innovation?
Across the supply chain, the US is the still most important player, so I wouldn’t make a blanket statement that in aggregate, the US is lagging behind, simply because it has a position that every other country envies, whether it be Taiwan or Japan or South Korea. But if you look at certain subsets of the chip industry, in particular logic manufacturing, the US’ competitive position relative to where it was five years ago is meaningfully behind and that I think is the trend line that has worried some US policymakers.
A decade ago, one can argue China’s biggest external vulnerability in this sector is its reliance on foreign imports of semiconductors. Would you say that over the past 10 to 15 years, China’s reliance on foreign imports has been greatly reduced?
No. In aggregate terms, China is buying even more chips from abroad. China’s capabilities have certainly improved, but it is still far, far behind the cutting edge and aggregate. If you want to make an advanced chip, you basically don’t have to use any Chinese inputs, whether it be software, materials or tools, and if you wanted to use only Chinese inputs, you couldn’t make anything anywhere close to the cutting edge.
If you look at revenue shares across the industry, China is also far behind the US. Even when compared to Taiwan or Korea, China is a smaller player. And where China is a player, it’s in pretty low tech segments in the industry. It has been making progress, and there are a few individual companies that are exceptions to that rule, but China’s competitive position in aggregate is really quite weak.
Do you think that since Taiwan has this monopoly in advanced semiconductor manufacturing, having this valuable asset is a meaningful deterrence to China’s potential plans of launching an attack on the island?
Maybe, but I’m far from confident in that dynamic. That would be true if it [were] generally true that economic interdependence dissuades countries from going to war, but it often doesn’t. 2022 has provided a frightening example of that, when Germany betted very heavily on energy interdependence with Russia, but ended up not getting a peaceful Russia.
I don’t think there is any reason for great confidence that technological economic interdependence alone is going to guarantee peace in the Taiwan straits. I hope it does, but it just seems far from obvious. For a long time people have clung to that explanation without actually subjecting it to serious questions about how valid it is.
Would you say relocating the hub of global semiconductor production from Taiwan to other places may actually have a negative impact on Taiwan and its security? Would that potentially expedite a potential military conflict in the region when Taiwan no longer has its ‘valuable asset’?
First off, I don’t think relocation is a plausible outcome. No one is talking about full relocation. TSMC is still investing most of its investment dollars and production in Taiwan, and it will remain a company that is mostly based in Taiwan for the foreseeable future, even with its construction of new facilities in other countries.
But, another way to ask the question might be, does the outside world or the US primarily care about Taiwan because of its advantages in semiconductors. The answer to that is obviously not because the US has committed to help defend Taiwan since before the first semiconductors existed, including during the second Taiwan Straits crisis. There are some people in Taiwan who argue it is only because of our semiconductors that we have a guarantee of our security, but that seems to me to be a complete misreading of history. The China-Taiwan political dispute is about a lot more than semiconductors.
Recently, the US Congress passed the CHIPS Act (CHIPS and Science Act). How big of a game changer do you think that legislation is?
I don’t think the CHIPS Act fundamentally shifts the playing field. It will lead to more chip manufacturing in the US than would have otherwise been the case, but not a dramatic shift in the geography of chip manufacturing. The big change is the bifurcation of the world’s tech sector into a China focused and a non-China focused situation. Every month that passes brings a new data point that suggests the dynamic of a more significant splitting of tech supply chains is increasingly visible and will continue over the next few years.
Is globalizing the chip industry a naive decision, and would you be criticizing globalization or advocating for economic protectionism if you agree with that position?
First off, I don’t think the chip industry is being globalized – it’s an industry dominated by four countries. I actually think the internationalization of the chip industry was a good thing, and I think the US has every reason to deepen its economic-technological operation with countries such as Japan and the Netherlands. In fact, any drive towards self-sufficiency is naive and will drive up costs without providing additional security.
However, there are two risks that I do worry about. One is the increasing risk of war in the Taiwan straits. I think almost everyone you talk to would agree with that statement. The second worry is if China’s role in the tech sector continues to grow, its ability to harness technology for military power will grow. What we find historically is that every country that has access to technology computing telecommunication systems inevitably deploys them for military and intelligence uses, the same with China. Given the state of the political contest between China and many of its neighbors, that to me is a worrisome development.
Your book Chip War won the FT McKinsey Book of the Year award in 2022. How do you feel about that and do you think you’ve gained a lot more attention and recognition from this? What can leaders expect from you next?
It was a great honor to get the prize – certainly unexpected from my end. As to what’s next; I’m still playing around with a few ideas, learning more about data centers and their political, intelligence and military ramifications, and also doing a fair amount of thinking about the combination of AI and Biotech and how that will lead to security risks in the future.