337: Computer Chips are Sexy and Profitable (Well Maybe not Sexy)
Buck: Welcome back to the show everyone. Today, my guest on Wealth Formula podcast is Chris Miller. Chris is an associate professor of international history at the Fletcher School of Law and Diplomacy at Tufts. Jean Patrick, visiting fellow at the American Enterprise Institute and Eurasia Director at the Foreign Policy Research Institute, is the author of The Chip War The Fight for the World’s Most Critical Technology. Welcome to the show, Chris.
Chris: Thanks for having me.
Buck: So I know the first thing we’re going to start basic here, right? Because I got a lot of, you know, smart people are in their fifties and sixties who’ve never really spent time trying to really understand what what the deal we know. We use computers every day. We have we know how powerful they are. And we hear the, you know, words for the you know, how much energy or how much how much power a computer has. But what exactly does a chip to?
Chris: A chip is a piece of material, usually a silicon that has thousands, millions or some cases billions of tiny electrical circuits carved into it. And those circuits are called transistors, and they are either on or off their on. They produce a one if they’re off, it produces zero. And these are the ones and zeros that undergird all software and all computing. So you can’t have memory without chips.
Buck: Got it. And you call the computer chips the most critical technology in the world. That is that right? That’s right.
Chris: Yeah. We’ve become so reliant on computer chips that our economy, our societies function without them. It’s not only computers or data centers that have chips inside. Your phone, for example, will have, in many cases, a dozen different chips. Not only the main processor that runs the operating system, but a chip managing the camera, one managing the Wi-Fi connection, another managing the wireless connection. Your car could have dozens and hundreds of chips inside your dishwasher, your microwave. Almost anything you touch with an on off switch has the chip inside. And that means that they’re crucially important to our everyday lives as well as to our economy. And that’s why it’s so dangerous that their production is all concentrated in Taiwan.
Buck: Got it. And when I guess one more background thing is, you talk about, you know, obviously the number of chips is important. Chips also change over time in terms of technology, too. Right. Can you tell us a little bit about, you know, how that how that evolves?
Chris: That’s right. Yeah. There’s a prediction known as Moore’s Law, which is not a lot, but this is just a projection going forward, which was made in 1965 and projected that the number of these tiny circuits that could be put on a chip would grow exponentially, double every two years or so. And since the 1960s, that has been borne out. And the result of that is that the amount of computing power on each chip doubles each year as well. So if you’ve got a chip that’s ten years old, it is far, far behind in terms of the amount of computing power it can produce. Then that’s the fundamental force that given us computers today in everything we use. That’s the reason why computers in the 1950s took up the size of us a large room, whereas today you can have more computing power than in that room in a tiny chip in your pocket.
Buck: Right. So there’s really two things. There’s the I guess the the you know, where the chips are going, who gets them. And also the technology involved as well. So these obviously we have so we have this background rate. We know we require chips for pretty much everything that we’re used to in modern society now. And I guess the question is, how do you you know, how do you relate that it’s almost like a commodity then, right, for the for the global economy? Is that is that fair?
Chris: In some ways that’s very it’s like a commodity in the sense that we can’t live without it. But it’s not like a commodity in the sense that producing them is extraordinarily complicated and a tiny number of companies monopolize their production. So oil, there’s lots of countries that have oil. The Saudis have more of it. But the Saudis produce ten or so percent of global oil production. Russia and the US something similar. By contrast, Taiwan produces 90% of the most advanced processor chips without which Apple or the company simply could not function. And then the machines used to produce. Chips are also monopolized by a small number of companies. For example, there’s one company in the Netherlands called ASML, which produces 100% of the world’s extreme ultraviolet photography machines, which are extraordinarily complicated rules about which is impossible to make a chip.
So it’s like commodities, but we need them. But it’s very unlike commodities in the sense that they’re monopolized by a small number of players and are more geographically concentrated than oil. Just to give you one other example, you know, Taiwan produces 90% of the world’s most advanced processors. OPEC, in aggregate, the Saudis, the countries, all of the members of all that produce less than 40% of the world’s oil. So this is much more vulnerability, much more half duration.
Buck: So along that lines, a stock about, you know, what dynamic that creates because obviously, you know, Taiwan is a is a is a place that sort of and the target of China. What I mean, I’m sure or maybe they’re not I don’t know but I’m sure U.S. government is quite cognizant of that fact. Why can’t we create more chip creating centers like why? Why can’t we bring that domestically? Or is it one of those things where it’s low cost labor and that’s what we’re what you need to to produce these things?
Chris: You know, it’s not primarily about labor costs because the chip making facility has some workers inside, but far more expensive other machines that are inside of it. And the most expensive of the machines inside a chipmaking facility can cost $150 million apiece. So it’s really a machinery cost that drives the cost of chip making. And to build one advanced chip manufacturing facility can cost $20 billion.
So these are the most expensive factories in human history. And so moving production from Taiwan or other centers of production like Korea elsewhere in the world is just brutally expensive. It would take hundreds of billions of dollars to fully move the industry away from East Asia towards the US. But what’s more is that it’s not a question of movement because the Taiwanese company that produces the world’s most advanced chips, the atmosphere, the Taiwan Semiconductor manufacturing company, is better than everyone else in the world. No one else can do what they can do on the island of Taiwan. So even if you had the funds to build a whole suite of new advanced chip making facilities in the US to make up for the capacity that produced in Taiwan, we still wouldn’t be able to do it as accurately and with as much precision as the Taiwanese can.
Buck: Just against a little background, why did this all concentrated on Taiwan and because am I wrong in saying that if you go back in the history of computing, that U.S. historically was the dominant producer initially?
Chris: That’s right. The first chips were invented in the United States in practice and also in in the towns south of Silicon Valley, San Francisco, that at the time wasn’t known as Silicon Valley, but now are there. Over the past couple of decades, production has shifted largely because for two reasons. One, governments in East Asia, especially in Taiwan and South Korea, and then now increasingly China have poured billions of dollars into incentivizing the shifting of manufacturing.
So that’s one big reason. Secondly, there’s a gentleman by the name of Morris Chang, who founded the Taiwan Semiconductor Manufacturing Company in 1987. And before that, he’d spent 30 years working in the US chip industry and we really at the at the pinnacle of his career when he moved to Taiwan and he had an insight that no one else in the US chip industry did, which was that the business model of making connections was shifting. In the past, most companies designed and manufactured chips in-house. The same company did both the design and the manufacturing. In 1987, Morris Chang founded TSMC on the Insight that if you only focused on manufacturing chips, you could serve a broad array of customers, manufacture a higher volume of chips, learn more from your higher production volumes, and therefore produce chips more efficiently and with more precision as a result.
And that’s been the business model of the Taiwan Semiconductor manufacturing company since that date. And that’s been why it’s been able to grow far faster than any competitor, whether in Korea and Japan, in the US, anywhere else in the world.
Buck: So you’ve set up a you know, essentially a geopolitical and ultimately and also macroeconomic scenario there. What’s happening at that level that in terms of, you know, China’s role, you’ve got China, you’ve got Taiwan. Tell us kind of what’s happening in that space right now. And you know, what scares you or what you’re encouraged about.
Chris: Or the situation right now is that the entire world is reliant on Taiwan, the US, Europe, Japan, China. Everyone needs the chips that only Taiwan can produce. Taiwan itself is hugely dependent on the chip making. Over 40% of Taiwan’s exports are semiconductors, and they largely go to China, where they’re assembled into servers, into computers, into smartphones that are then reexported worldwide. Both the US and the Chinese military are dependent on chips from TSMC, the US military can access them legally. The Chinese military, we know from any study, has ways for accessing those chips illegally and as a result, both the US and the Chinese military are building up their capabilities to fight over Taiwan with the chips produced by the entity.
The US is trying to change this by further restricting the types of chips that can be sent to China and Vietnam is subject to many US export control rules. So over time China will have less access to the chips that Vietnam produces. But that creates a different type of risk, which is that China will try to gain access, or at least to cut the US off by attacking Taiwan. And if there were to be a war on the Taiwan Straits, the disruption to the chip industry on an island would be almost certainly complete, in which case we’d find it impossible to build a smartphone anywhere in the world the next year to say nothing of PCs or telecoms infrastructure or data centers. The cost would be enormous measured in the trillions of dollars. I think worse probably than the economic cost of COVID and the subsequent lockdowns.
Buck: Yeah, that’s like a big deal, right? I mean, could the US remain the number one economy in the world if it doesn’t domestically dominate the chip market? Eventually?
Chris: Well, it’s a question mark. I think in the past we’ve seen that many of the advances across society, not only in the tech sector, but more broadly, have been driven by computing power, whether it’s in medicine, whether it’s in automobiles, whether it’s in consumer goods. Computing power is crucial. And computing power means chips and if you think about actually next generation technology, you take the cars of five years from now versus the cars of five years ago. They’re going to be even more reliant on computing and sensing and memory. And that means chips. And if you contrast that Tesla to a lot of cars today, what you’ll find is Teslas have more chips, more than chips in them as a result. So I think it’s hard to imagine that there won’t continue to be a really close correlation between the amount of computing power you can produce and get access to and the fate of your economy more generally.
Buck: There was a there was the Chips Act, right, which was, I guess a $53 billion direct investment. Tell us about the impact on the US chip sector there. Is that is that just not enough, is it? I mean what.
Chris: That’s right. It’s just not enough. That’ll by you to chip making facilities more or less. It’s a good start. I think it’s the right move. It’s a recognition that we’ve become dangerously, horrifyingly reliant on Taiwan, but it’s only barely going to move the needle. And I think there’s more to be done, both in terms of reducing the cost gap of producing in the US versus producing in other countries and pushing back against unfair subsidies that other governments are giving their chip industries and then also in making sure that US technology is not leaking abroad and being used to support, in particular the Chinese military, where we know the Chinese military has pretty widespread access to advanced US, designed US made chips.
Buck: Right. And along that lines too, I mean, I guess the question is, is it an argument to advocate for subsidy zation of a US chip market production so that we can encourage this? And are you seeing any signs of that?
Chris: You know, I think it’s an interesting it’s an interesting, complicated balance that must be struck. On the one hand, there’s no doubt that Chip making over the past 60 years has been driven forward by private actors far more successfully than the government. Governments played a role in funding R&D and funding science, but it’s been companies that have really found ways to produce chips the most efficiently.
So it would be wrong, I think, for us to want the government to start playing a much bigger role in the industry. But so long as every other government in the world is providing incentives to manufacture in their countries, the US can’t simply fall behind. And the reality is that everyone else in the world Europe, Japan, Singapore, Taiwan, Korea and most importantly, China is pouring money into the Chinese are pouring not only billions but probably hundreds of billions of dollars over this decade into the domestic chip industry, which is going to have a really profound effect if we don’t do anything in response to that. We’re not only going to face a situation where the chip industry concentrated in Taiwan as it is today, but also where we’re going to be even more reliant on chips made in China as well.
Buck: Right. So so I guess, you know, haven’t had an opportunity to read your book, but I’m guessing a big part of the fight for the world’s most critical technology, as you’ve called it, is is really to ramp up, you know, our our you know, our sense of importance of this issue. Is it your sense right now that the U.S. under the U.S. right now is underestimating this problem and is
anything, you know, positive coming down the line that might change that?
Chris: Yeah, I think it’s clear the US is underestimating the problem. And part of the reason why I called the book Chip War is that this isn’t simply a sort of hypothetical technological question or something that only matters for the speed at which your computer operates. This is something that has profound military applications as well, because when you think of what make military systems work today, it’s in many cases the electronics inside. And so, for example, when we’ve taken apart the guidance computers of Russian missiles that have fallen on battlefields in Ukraine, what you find is that they’re often full of foreign made chips us, South Korean, Taiwanese, because the Russian domestically produced chips are so bad that the Russians find it in their interest to smuggle in chips they’ve bought off the gray market to put in the missile systems.
Now, China doesn’t have to do that nearly to the same degree because they’ve got a much more advanced chip industry already. But when both the US military and the Chinese military think about the future of war, autonomous drones, for example, or advanced electronic warfare systems, this is all about semiconductors processing power, memory capacity, digital signal processing. These are all, all only possible with chips. And the better chips you have, the more processing power is available. So it’s it’s a chip in a literal sense too, in that whoever’s got access to the most advanced chips will have better military power, more successful battlefield weaponry. And that’s that’s why this is a real chip war. And that’s why I titled the book Chip or Self.
Buck: And how do you invest in chips? Say I’m a, you know, like we have a bunch of people listening to this podcast thinking, well, gosh, yeah, this is this is actually a big issue. And the chips are know if I’m thinking about them as a commodity, I mean, it’s a lot more valuable than gold right now. I mean, it’s so what I’m not talking about specific companies or stocks, but if you look into this area, like how do you how do you get exposure to that? If you’re an investor?
Chris: Well, there are a couple different types of companies in the chip ecosystem. There are companies that design chips. There are companies that manufacture the chips, and then there are the companies that produce the tools with which you you can’t you need to make chips and you can’t make chips without them. And in the tool making, there’s a lot of companies that are basically monopolies in their position. And so what that means is that they’re much less vulnerable to competition and much less susceptible. They still are susceptible to market movements, whereas the company that design chips and some of the companies that make chips fluctuate much more with consumer cycles. And so right now what we’ve seen is after a couple of years of pretty high valuations and in chip equities, we’ve seen them draw down because there’s an expectation of lower consumer spending across the world and therefore fewer smartphones and things like that being bought, which means fewer chips needed.
But if you look at the tool makers that are actually making the ultra precise machinery behind this, they’ve seen less fluctuation because demand for their tools is still very strong and their tools will be used not just for years, but for decades and chip making facilities.
Buck: And and so the book again is called Chip War The Fight for the World’s Most Critical Technology. If there are any other messages from this book that you think are, you know, worth highlighting that we haven’t spoken about.
Chris: I think the final thing I’d say is that you can’t understand trade or globalization. I think that’s the conductors. I mentioned that for Taiwan, chips were 40% of their exports. But what most people don’t realize is that for China, chips are no less important for understanding their trade flows. China spends more money each year importing chips than it spends importing oil. And so if you’re thinking about the Chinese economy, you’ve got to consider the impact of chips because they’re one of the most important inputs into China’s balance of trade. And so when you look at global trade flows, the global economy, we think a lot about oil, we think a lot about commodity prices. But actually, in many cases, chips are even more important and input into the stuff of economies than commodities we focus on much more frequently.
Buck: Interesting stuff, Chris. Again, the book again is Chip War The Fight for the World’s Most Critical Technology. I’m sure it’s available everywhere. Amazon and the usual other outlets.
Chris: It is on Amazon as of yesterday.
Buck: Okay, great, great. And you also have a website, ChristopherMiller.net, what kinds of information we find there?
Chris: Well, I’ve got articles, reviews, my book, public events on the website and mostly focused on the new book Chippewa.
Buck: Yeah, fantastic. Well, Chris, thanks again for being on Wealth Formula podcast and I love to have you on again at some time.
Chris: Great. Thanks for the invitation.
Buck: We’ll be right back.