Interview: Chris Miller, historian and author of “Chip War”
In which I consult an expert on the battle to control the semiconductor industry.
(Noah Smith’s interview with Chris Miller.)
The struggle to control the semiconductor industry is one of the most important economic stories in the world today. Whether China can wrest dominance of semiconductors away from the U.S. and its democratic allies, as it has so many other high-tech industries, will go a long way toward determining the military balance of power this century. And the best book you can read to familiarize yourself with the basics of this titanic struggle is Chip War: The Fight for the World’s Most Critical Technology, by Tufts University historian Chris Miller. I reviewed the book here; its place at the top of many lists of the “best books of 2022” was well-deserved. I’m not sure I’ve ever read a book that does such a good job of crafting key events into a tight, readable narrative while also teaching readers key facts about a complex technology.
But what’s even more amazing about Chip War is that the book came out just a few days before the Biden administration launched a sweeping regime of export controlsaimed at stifling China’s high-end chip industry. Naturally, Chris Miller has been in high demand since that event, since he’s the expert who’s best positioned to explain those controls to the public, and to predict what they might accomplish. In this email interview, I asked Miller a wide-ranging series of questions about the export controls, China’s efforts at retaliation, the CHIPS Act, the U.S.’ need for semiconductor workers, Japan’s attempts to revive its own chip industry, and more. I learned a lot. Enjoy!
N.S.: My first burning question when it comes to semiconductor export controls is: Why now? I have to say, I was personally pretty surprised that the U.S. went ahead and just basically declared economic war on the Chinese chip industry; I had expected us to use the possibility of those controls as a threat to deter China from making moves against Taiwan, but we just went ahead and did it. What prompted us to make that big move?
C.M.: I was less surprised, to be honest, for three reasons. First, the U.S. government is very concerned about Chinese intentions over Taiwan over the next decade. There are only a limited number of tools the U.S. has to influence Chinese military and intelligence capabilities over a 5-10 year time horizon, but trying to slow China’s chip industry and its compute capabilities are one plausible way to do it, especially as defense and security establishments around the world experiment with how to apply AI to military systems. Second, China has been catching up in certain parts of the chip industry. The picture is mixed, but certain segments (eg YMTC, a memory chip firm) were making impressive strides while using tools imported from the U.S. and allies. So it wasn’t plausible to claim that, assuming the status quo persisted, China wouldn’t keep making progress. Third, the argument for waiting — for using the threat of export controls as a deterrent — didn’t seem compelling, especially given the failure of U.S. sanctions to deter Russia from attacking Ukraine.
N.S.: That all makes sense. How much impact will the export controls have on China’s military capabilities over the next 10 years? I’ve heard it said that military tech generally uses trailing-edge chips; if so, that would mean that in the short term, China’s military would only need chips that China can already make, using tools they already have. How true is that?
C.M.: Autos provide a good analogy for understanding how militaries use chips. A typical new car might have a thousand chips inside. Most are very simple, like the ones that make your window move up or down. But the high-value features–the entertainment system, the lidar or radar sensors, and the semi-autonomous-driving features, all require more sophisticated and specialized semiconductors. What’s more, a lot of the high-value features in cars don’t only require chips on cars–they also require sophisticated chips in cell towers and datacenters too. This is why Tesla builds its own high-end Dojo chips.
Military systems are pretty similar. Most of the chips in tanks and missiles are low-end, but the chips that provide differentiated capabilities are not. Just like autos, some of the most sophisticated chips aren’t actually in the missiles and tanks, but in the networks and datacenters that guide and train them. We know that autonomous driving efforts require huge volumes of advanced chips in cutting edge datacenters. We know less about the U.S. military’s drone programs, but there’s no doubt they use a lot of sensors, a lot of communications, and a lot of compute. The Himars missiles used in Ukraine don’t require ultra-advanced chips themselves, but they rely on targeting information provided by a vast array of sensors and processors to sort signals from noise or to differentiate tanks from trucks. It’s now easy to put GPS guidance in a missile, since every smartphone has GPS guidance too. But can your missile maneuver itself to avoid countermeasures while operating in an area where GPS is jammed? If so, its going to need more sophisticated semiconductors.
There’s not a single type of chip for which you can say “without this chip, China’s military modernization will grind to a halt.” It’s always possible to design around a certain component. But the more you have to design around subpar semiconductors, the more tradeoffs you have to make between performance, power consumption, reliability, and other characteristics. I think the recent tightening of export controls will exacerbate these tradeoffs.
N.S.: So the goal is really just to slow China down, keep them half a step behind us. That brings me to probably the most important argument against export controls. A lot of people argue that had the U.S. not enacted export controls, China would have remained dependent on U.S. chips for longer, but now that we cut them off, China will simply learn how to make everything itself, thus cutting U.S. companies out of the market and ultimately raising China’s own technological capabilities. What do you think of this argument?
C.M.: I think its hard to sustain the argument that the controls will make China pursue a strategy of reducing dependence on the U.S…because that was already China’s strategy. Chinese leaders, including Xi personally, have articulated this repeatedly since at least 2014. They launched a major industrial policy program focused on the aim of ending reliance on the U.S., spending billions of dollars annually. So to say the export controls caused this goal gets the chronology backward: this goal existed for years before the export controls.
Now, one could argue “China’s prior policies weren’t working and reducing dependence on the U.S., but now China will pursue more effective policies.” But I haven’t seen anyone articulate why this would be the case. It doesn’t seem like semiconductor funding in China has increased (and the sums involved were already vast.) Nor have the export controls introduced new information into the Chinese policymaking apparatus that will make it smarter. Beijing was pursuing this self-sufficiency strategy before the controls precisely because it knew it was so dependent.
Perhaps you could argue that the imposition of controls has reshaped the political economy or the relationships between Chinese firms and government in a way that will lead to smarter Chinese policy. I haven’t seen anyone spell out how this might work. So I’m skeptical, and I think loss of access to chipmaking tools and the broader chilling effects on expertise transfer will make China’s catch up efforts harder.
N.S.: How difficult will it be for China to make chipmaking tools similar to those made by ASML? I know they’re trying very hard to steal ASML’s tech, and I’ve seen one report indicating they may have had some success there. Also I’d expect them to try to purchase ASML machines through third countries, as well as accelerating their own indigenous R&D efforts. Will any of those workarounds work, and if so, how long until they catch up?
C.M.: The likelihood of purchasing these machines through third countries is close to zero. The number of advanced tools produced each year measures in the dozens, and there are only a handful of customers. A single advanced lithography machine requires multiple airplanes to transport. And there are ASML staff on site at all times who are critical to its operation. So its difficult to imagine a set of tools that would be more difficult to smuggle.
Replicating them is easier, but still a monumentally challenging task. It took ASML three decades to develop EUV lithography tools, and it was only possible in close collaboration with users like TSMC and Intel. Of course, it will be easier to replicate the tools than it was for ASML to first produce them. But these are the most complex and precise pieces of equipment humans have ever made. The challenge isn’t only to replicate the unique components inside the tools – such as the smoothest mirrors humans have ever made – though this will be hard. The really challenging part will be to get the hundreds of thousands of components to work often enough so that the tools can actually function in high-volume manufacturing. If each of the hundreds of thousands of components in your tool breaks down once a year, the tool basically never works. So reliability is a potentially fatal challenge.
And remember–lithography tools are probably the hardest challenge, but they’re not the only one. There are also deposition tools, etching tools, metrology tools, and others. China is behind to varying degrees–often significantly–in all of them. All these tools require tens of thousands of precision components and need to be accurate at the nanometer scale.
The final point here is that all the Western toolmakers have new chipmaking equipment rolling out on a regular basis. ASML will soon release its next generation lithography tool, called high-numerical aperture EUV. The industry continues to race forward. So if China manages to produce its own suite of EUV lithography and related etch, deposition, and lithography tools within five years, it will still be substantially behind the cutting edge.
N.S.: Let’s talk a bit about the U.S.’ own efforts to rebuild domestic chip fabrication capability. One very common complaint, from both TSMC and domestic chipmakers, is that building and operating fabs in the U.S. costs much more than in Taiwan or other developed Asian countries. How true is that, and what are the sources of the excess costs?
C.M.: For a cutting edge fab, around 70% of the total cost is in the chipmaking tools, which cost the same everywhere in the world. A critical factor in cost differential is tax policy around this capital expenditure. Because the capital investment involved in a chip fab is so massive, tax treatment of capex and depreciation schedules are hugely important. Sounds arcane, but its a many-billion-dollar question.
Public debate about the CHIPS Act has mostly missed the fact that, alongside the incentives for chipmaking (the money the Commerce Department is distributing), the act also includes a very generous investment tax credit that will be just as significant in terms of financial impact. However this tax credit expires at the end of 2026. Whether it is extended will have a major impact on the cost competitiveness of chipmaking in the U.S. over the long run.
The other sources of cost differential are more challenging to deal with. There are some labor cost differentials in fab operation, but chipmaking engineers are paid reasonably well in Taiwan and Korea, so this factor isn’t critical. Construction costs are substantially higher in the U.S. than in Taiwan, for example. Power, water, land, and other inputs are also sources of cost differential. Regulations are a major problem in the U.S. Due to excessive permit requirements, largely imposed by NEPA, it takes longer to build a fab in the U.S. than in Europe–a jurisdiction not known for light-touch regulation. The U.S. hasn’t done nearly enough on this front.
Finally, ecosystem effects also matter. There’s a lot of niche know-how involved in building fabs. The Taiwanese and Koreans have built a lot of fabs recently, so they’ve honed this know-how, both at chip companies but also at the large network of suppliers they rely on. As fab construction in the U.S. slowed in recent years, some of this expertise deteriorated. I suspect the current fab construction boom will rebuild some of this know-how and drive down this source of cost differential.
N.S.: I’ve heard the fabs that TSMC and Samsung build in the U.S. will be mostly for trailing-edge chips, while they keep leading-edge production in their home countries. Is that true? And if so, does that mean U.S. leading-edge production will depend on Intel catching back up with its rivals? How likely is that?
C.M.: TSMC has said that its Arizona facilities will produce one generation behind what is manufactured in Taiwan. So that’s still pretty close to cutting-edge, though it isn’t the most advanced chips. This means iPhone processors won’t be made in Arizona, but that other types of advanced chips for datacenters, autos, and other uses might be.
You’re right, though, that Intel’s future will have a major impact on the U.S. chipmaking landscape. Its an American firm with most of its R&D in the U.S., and most of its manufacturing in the U.S. or Europe–so its success or failure will shape the scale of advanced chipmaking investment in the U.S. Intel’s manufacturing process technology fell behind TSMC’s the past few years, though it has a plan to catch up to TSMC by 2025, and over the past year or so it has hit every technical target along this catch-up path that it has set.
N.S.: I’ve noticed that Intel originally succeeded not by outcompeting foreign companies in the memory chip market, but by innovating a new market in the form of microprocessors. Now NVidia has become hyper-successful not by competing with established players, but again by innovating in a new market — this time in GPUs for AI. This makes me wonder whether U.S. chip industry promotion efforts should focus more resources on supporting startups rather than big incumbents. What do you think should be the balance there?
C.M.: I think we need both. The current level of geographic concentration in chip fabrication and assembly is dangerous, and no number of startups can address this in the short run. If a crisis breaks out in East Asia, we’ll look like Germany and Nord Stream, but ten times as naive because the economic impact will be vastly worse.
But I agree that in the longer run, the chip industry will be defined by its most innovative firms, so trying to help incumbents isn’t a great long-run strategy. All of the manufacturing incentives in the CHIPS funds will inevitably go to incumbents, because there aren’t any startups when it comes to high volume manufacturing. But I think the R&D funds in the CHIPS Act should be structured to help startups. I do worry about the risk that big firms capture the R&D programs, too. They certainly shouldn’t be excluded, but they’re capable of funding their own R&D and already spend many billions of dollars annually. We’re better off using R&D funds to facilitate startup formation, e.g. by giving startups access to expensive tools that they otherwise wouldn’t be able to afford.
N.S.: One thing you didn’t mention yet, which a lot of people talk about, is talent and labor for U.S. chip manufacturing. Do we need to be bringing in a ton of workers from overseas, and where should we be bringing them from?
C.M.: I think we need both to be building up training pipelines for U.S. workers and simultaneously increasing pathways for high-skilled immigration. We’re actually making significant progress in training, with a number of new programs having been set up by chipmakers and universities. I think the CHIPS Act should have been coupled with a CHIPS Visa program to make sure companies have fast-track access to the workers they need, though the Commerce Department is trying to take steps to facilitate visas for semiconductor workers within the boundaries of existing legislation.
And if you’re really interested in strategic competition with China, a visa program specifically targeted at semiconductor workers from China would be a win. America’s ability to strip its adversaries (USSR, Iran, China etc) of their brightest minds has been key to its geopolitical success.
N.S.: If you were advising the Biden administration, what would you list as the top action items or priorities to improve the U.S.’ position in the semiconductor industry, beyond what has already been done? Also, by the way, are you advising the Biden administration on this?
C.M.: In the short run, there’s more work to be done on making the U.S. cost competitive. I mentioned permitting reform. We should recognize Korea’s and Taiwan’s safety and construction regulations for fabs as equivalent, so that firms from those countries don’t need to redesign their facilities when they want to build in the U.S. The more they can copy and paste from what works in those countries, the less money they have to spend redesigning facilities to suit the needs of America’s fire and plumbing inspectors, who have much less experience with fab safety than the biggest firms. (Moreover, with billions of dollars of equipment in their fabs, chipmakers have plenty of incentive to avoid accidents.) Second, there should be strict time limits in which permits are either rejected or approved, so that the NEPA burden can be limited. At the very least we should be able to make our regulations only as burdensome as Europe’s. Today they’re worse.
The second short-run change is to extend the investment tax credit, which currently expires at the end of 2026. It should be made permanent to ensure that manufacturing in other countries isn’t cheaper simply for tax reasons.
In the long run, whichever country innovates most rapidly will succeed. The CHIPS Act puts more money into R&D, and there’s discussion of focusing CHIPS funding toward prototyping rather than basic science (which is great, but which we already have plenty of.) In the chip industry, prototyping is very expensive, so we have fewer startups and new products than, say, in software, simply due to the high upfront cost. Making it cheaper and easier to turn new ideas into working prototypes like test chips would help boost the rate of innovation.
N.S.: One news story I’ve seen a number of Americans express confusion about is the Micron saga. China’s government recently banned Micron Technology, a U.S. company that makes memory chips, from selling its chips to many companies in China. This was widely seen as retaliation for the U.S.’ export controls. U.S. Commerce Secretary Gina Raimondo said that the U.S. won’t tolerate the ban. But how does this make sense? If the U.S. is banning China from buying some of our chips, how can we get mad when they refuse to buy even more of them? And to make matters more complicated, Micron now appears to be pumping huge amounts of money into a factory expansion in China. Why are they doing that, and should we be worried about it?
C.M.: The U.S. understandably wants China to keep buying as many U.S. chips as possible, both for the sake of American firms, but also because it means that China’s chip spending is not helping to fund Chinese R&D. Yet we should probably assume that China will buy fewer foreign chips over time. After all, this has been the primary goal of China’s industrial policies like Made in China 2025. Xi Jinping has repeatedly articulated the goal of buying fewer foreign manufactured goods, and in many ways it has succeeded. The US has understandably opposed this aim as contradicting the logic of international trade. But the desire not to buy foreign manufactured goods is hardwired into the way China’s leaders think and the way the system acts. So we shouldn’t be surprised to see China continuing to pursue this goal.
N.S.: Also, I’d like to ask about Japan. Why has their semiconductor industry fallen off so badly? And what can they do to get back in the game?
C.M.: Japanese firms made two key errors in the late 1980s and early 1990s, when it seemed like Japanese chip firms were taking over the industry. First, they focused on market share rather than profitability. They won huge market share, but because hardly any of them made money, this wasn’t sustainable. This is an error that China is replicating today.
The second error Japanese firms made was to be insufficiently focused on software. It sounds counterintuitive, but many of the best hardware firms have succeeded because of software. Intel’s x86 chip architecture became industry standard for PCs because of its integration with Windows. Apple has spent so much money designing its own smartphone chips so that they work seamlessly with iOS. Nvidia makes top-notch GPU chips, but what really differentiates its GPUs is that everyone in the industry is familiar with the programming interface, CUDA, that the company spent years and billions of dollars developing.
None of the Japanese chipmakers developed a type of chip with a moat created by its software interface. They ended up mostly competing in the market for memory chips, which is a commodity sector in which price matters above all else. It’s a bad niche for a country with higher costs than all of its neighbors. U.S. firms realized it is far better to be in market niches where a unique software interface shields you from head-on competition and lets you differentiate more easily on quality.
N.S.: How much should the U.S. governments coordinate with the governments of Japan, South Korea, and Taiwan in order to create a robust international semiconductor industry and supply chain that doesn’t depend on China? Should we view chip production in our Asian allies as a pure win, or should there be some intra-alliance competition there?
C.M.: There’s a balance to be struck. We don’t want governments to think they can collectively manage the chip industry. They can’t–or at least they won’t do it well. I get nervous when political leaders use phrases like “supply chain” (which they don’t understand in sufficient detail) or “resilience” (which doesn’t mean anything.) But we need international cooperation to address the two aspects of China risk related to semiconductors–the risk that China disrupts our access to chips from Taiwan and the risk that China catches up technologically. Neither of these problems can be solved by the U.S. acting alone.
N.S.: Should we follow up current export controls on China with new successive rounds, in order to keep them on the back foot? And if so, what should future measures look like?
C.M.: Export controls are inevitably a game of whac-a-mole, as China tries to set up shell companies. So the controls will have to be updated regularly. However, I’d be cautious about major new expansions in the controls, at least for now. First, because I think the controls are likely to work at delaying China’s progress toward the most cutting edge capabilities. Second, because expanding controls to lower technology levels–where China already has greater domestic capabilities–is less likely to work because China has more domestic capability to replace U.S. tools at lower tech levels. And third because, right now, U.S. policy is in the export control sweet spot: where controls are likely limiting China’s technological progress, but where China still believes its best strategy is to spend many billions of dollars each year buying U.S. chips and thereby funding U.S. chip firms.
N.S.: What are some quantitative metrics we should be keeping an eye on in the semiconductor industry, in order to know how the international competition is going?
C.M.: In terms of technological leadership in the chip industry, a key question will be at what rate leading Chinese firms advance their manufacturing processes and how this compares with non-Chinese firms.
But I think the more pressing short-term metric is market share in China’s domestic chip market. Today China’s domestic chip market is dominated by foreign firms. China’s leaders have repeatedly stated they want to change this by importing fewer. That’s the point of Made in China 2025 and other industrial policy plans. I wonder whether they might finally take steps in this direction — not by overtaking competitors technologically but by pressuring Chinese buyers to use less capable domestically produced chips.
The electronics industry is the only major sector of the Chinese economy that has not thus far been subject to substantial “buy Chinese” pressure. (In contrast to autos, aviation, high speed rail etc.) In most other sectors, “buy Chinese” has been an acceptable policy because Chinese firms learned to produce products at or close to the technological frontier. Could we be at a point where China’s leaders are so committed to self-sufficiency, they decide to pressure domestic firms to buy domestic chips, even if they’re worse? The implications for global trade would be dramatic, because China spends as much money importing chips as anything else.
N.S.: Final question. If you had to predict how the U.S. chip industry will look in 2030, what would be your general predictions?
C.M.: I would make two predictions. First, we’re going to have more bifurcation between the Chinese sphere and the non-Chinese sphere. International media focuses on U.S. policy, but I think current and future Chinese policy is potentially even more important in driving bifurcation. Second, I think the rate of improvements predicted by Moore’s Law will continue, which means that chips in 2030 will be 8x as powerful as they are today. It will be hard to find any other segment of the economy that will come close to this rate of productivity improvement.
(This post is republished from Noahpinion.)