The semiconductor is among the few inventions that has fundamentally changed our civilisation. In the last seven decades, the microchip industry has enabled devices to get smarter, process large amount of data, run multiple tasks, and enhance missile system’s strike capabilities. Today, chipsets power smartphones, computers, TVs, washing machines and various other consumer durables. The story of how these microprocessors found their way into almost all of our modern-day devices is nothing short of a masterpiece. But tracking the people driving innovation in the industry, advances in chip-making processes, and geopolitics surrounding technology transfers is no mean feat. Chris Miller’s Chip War , which has been named The Financial Times ’best Business Book of the year, offers a panoramic view of the semiconductor industry and reveals how the pivotal component has become the lynchpin for global dominance. Miller spoke to The Hindu ’s John Xavier on semiconductors, government policies to support chip-making locally, the complexity of global supply chains, technology transfers and geopolitics, and the issue of Taiwan. Edited excerpts.
You covered roughly 75 years of history and geopolitics in this book. How did you approach it?
Chris Miller: When I started writing the book, I wasn’t planning to write on semiconductors. My initial interest was in the development of missile technology during the Cold War. As I came to learn more about how missiles have changed over time, I realised the guidance computers inside were the interesting part of what has changed. That led me to realise that semiconductors were a critical component of military power. I hadn’t known this before I started the research. Thus, a book that was intended to be about military technology ended up being about something much broader as I learned more about how the technology had developed.
Can you elaborate that process? Were you specifically looking at Cold War or you already had China in mind? How did that play out?
CM: I was looking at the Cold War as a primary focus for the book. But I was doing that in 2016-17 when the U.S.-China tensions were beginning to escalate. China was still dependent on U.S. and Taiwanese chip technology. I realised that there was a link between the Cold War story and the U.S.-China dynamics today. The third point that convinced me to write the book was when I learned that China spends as much money importing chips each year as it spends importing oil. I understood then that the image of globalisation that I’d had in my mind was missing something fundamental because you cannot understand global trade without putting chips at the absolute centre of the story. Then it became clear that the last 50-75 years of military history could not be understood without semiconductors, neither the U.S.-China relationship. As I learned more about it, I came to the conclusion that it was time for a book that put it all together, and explain why semiconductors are critical for all of these key trends shaping the modern world.
Most governments are now offering subsidies to make microchips locally. You write how federal support could not help GCA, a top-tier semiconductor player. Through the GCA’s experience, what is your view of the Chips Act and other performance-related incentive schemes offered by governments?
CM: Everything depends on how the support is structured as no government in the world has enough money to match what private industry is willing to spend in the chip industry. The best way to illustrate this is by looking at the U.S. government’s in-house semiconductor manufacturing for security services. Two decades ago, the intelligence agencies decided they simply couldn’t produce chips in-house.
Currently, a new chip making facility can cost $20billion-25 billion, which is twice the cost of an aircraft carrier. But an aircraft carrier is in service for about 50 years, whereas a facility for making cutting-edge chipsets lasts a couple years due to advances in the manufacturing process. So, governments can’t do it on their own. But they can find ways to be supportive of private industry. But if they try to replace private industry, that is a hopeless task.
So governments must look at the broader scheme of things when attempting to let companies invest locally?
CM: Yes, one single government can’t support the industry as this is a very complex set of things coming together. The companies that governments support eye the consumer market to sell their products. Governments alone cannot provide the market. Can they provide some support at the start-up phase? Possibly. But ultimately, they have to find the end market. For almost all chip firms, the end market is some sort of consumer device.
Government programmes in Japan or India must help companies find big markets in the consumer sector. Unless they think of ways to find large markets of substantial volume, there is no way to make chip production commercially viable. You got to have a lot of volume because of the economies of scale in the industry, and governments alone cannot provide that volume.
This leads us to another important aspect of this industry, the supply chain. What is your view of the current semiconductor supply chain?
CM: The semiconductor shortages in the past couple of years have induced many people to think that the supply chain is fragile. But the past 20-25 years suggest the opposite; it survived the tsunami in Japan that knocked out several chip-making facilities, earthquakes in Taiwan in the 1990s. Thus far, many of these supply disruptions haven’t really caused major shocks to the industry. The industry has actually been quite resilient to all sorts of natural disasters. The shortages during the pandemic of the past two years were driven not by a problem in supply but by a surge in chip demand. If you look at the production of chips globally, it increased by 8% in 2020 and then by double digit rates in 2021. The problem was just that demand grew by even more than supply. This is a story of a supply chain that’s pretty resilient both to economic surprises and natural disasters.
What is the biggest challenge?
The biggest challenge the chip industry faces is that we haven’t yet had a test of the industry’s resilience in the face of a geopolitical disruption. The reality is that the most plausible large-scale geopolitical disruption, which is something going wrong in the Taiwan Straits, would happen at the absolute epicentre of the world’s chip production in Taiwan and to a lesser extent in China. Although the chip supply chain has been resilient to economic and natural disasters, it is completely unprepared for the coming geopolitical shock -- a possible Chinese attack on Taiwan.
Why is Taiwan at the heart of it?
If you look at the numbers, 90% of the most advanced processor chips are produced in Taiwan, and over a third of the processing power added annually comes from Taiwan. There’s just no way the world can make up to that in a short amount of time if it loses access to Taiwan. The reality is that we are so reliant on chips today that a reduction in supply would absolutely destroy the global economy. It could lead to a 1929-style Great Depression as manufacturing output will drop in almost every country. The risk is that a Chinese attack in Taiwan is getting more likely every year.
But you point out that it is in the best interests of China to not harm Taiwan.
Two decades ago, it was clear what the military balance of power was in Asia -- that the U.S. was a dominant power, and China was by far the weaker power. Today that is not clear. In fact, it is very uncertain which power has the most capabilities. That uncertainty creates a risk. It also creates an uncertainty on whether Chinese policymakers might decide that they have got a real chance of accomplishing a goal they have long held since 1949 -- taking control of Taiwan. The challenge is that there’s no scenario in the next couple of years in which that uncertainty about the military balance is going to change. It’s only going to get more uncertain because if you look at Chinese military expansion plans over the next several years, they are going to be building dramatically more ships, planes and missiles than the U.S.
You make an interesting forecast on China: “If the trends of the late 2010s were projected forward, by 2030 China’s chip industry might rival Silicon Valley for influence. This wouldn’t simply disrupt tech firms and trade flows. It would also reset the balance of military power.” How does this forecast hold?
CM: It’s hard to have any certainty as to whether China will be able to develop the capabilities in 10 years’ time to rival Silicon Valley. They certainly got a shot at developing these capabilities. A lot depends on the efficacy of their government support for their industry and a lot depends on decisions made by countries like Korea, Japan, the Netherlands, which could provide a lot of important technology to China. Right now, there’s a big debate going on among key Western powers as to whether there should be a multi-lateral system of tightening tax controls on China or not. The U.S. has unilaterally moved, but it doesn’t control the key technology. Japan and the Netherlands will be important players. If China is able to access advanced technology from Japan, the Netherlands, Korea, and others, then it will have a much better chance at catching up.
Whatever China’s success in catching up is, that success will feed into military power because when you look at the types of advanced systems that defence planners both in the Pentagon and in Beijing are planning to field in the next 10-20 years, they are going to be more reliant on computing power, secure communications, advanced sensors and digital signals processing. All of which means more reliance on advanced semiconductors.
There is going to be a direct relationship between a country’s ability to access advanced semiconductors and its ability to field the types of increasingly autonomous, data-driven military systems that all of the world’s defence planners believe will define the future of military power.
Staying within geopolitics and on building allies, what about the dispute between Qualcomm and Arm over Nuvia’s licences from the latter? How do you view such legal battles?
CM: It’s a tricky dynamic. If the focus is solely on retaining the current structure of the chip industry to contain China, that will have the effect of slowing innovation and make it harder to contain China’s tech advances in the long run. What we need to do is find ways to both restrict the transfer of sensitive technology to China, but also to retain the competitiveness of our [Western] economies and our [Western] tech sectors. That’s a balance that will be complex to strike in some cases, especially given that the export control-oriented officials and governments like the fact that they got monopolies where officials focused on competitiveness don’t like their monopolies. That is a tricky balance to manage. I don’t think we should move in a direction of fully embracing monopoly positions at the expense of innovation because ultimately what really made the chip industry hard to replicate over time is that it races ahead in line with Moore’s Law and we have got to keep it going as long as possible.
Last November, Taiwan made a diplomatic move in Lithuania. Where do you place this EU nation in the overall scheme of things in the context of China, U.S., and Taiwan?
CM: China-Taiwan is a complex situation because when most other countries look at the concentration of chip-making in Taiwan, they get very nervous. But when Taiwan looks at it, it sees it as the guarantee of future economic success. And so, Taiwan is trying to find a way to make allies like the U.S., Japan and other countries comfortable with the geographic distribution of chip making. So, it will open up new facilities outside of Taiwan. It will also try not to have a situation where Taiwan’s leading company ends up outside of country because that would be problematic for the Taiwanese economy. The Taiwanese government is doing a reasonable job at it right now, but it is a balancing act.