“World War II was decided by steel and aluminum, and followed shortly thereafter by the Cold War, which was defined by atomic weapons,” Chris Miller, a professor at Tufts University’s Fletcher School of Law and Diplomacy, writes in the introduction to his latest book. So what’s next? According to Miller, the next era, including the rivalry between the U.S. and China, is all about computing power.
That tech rivalry and the story of how the chip industry got from four to 11.8 billion transistors are all part of Miller’s book, “Chip War: The Fight for the World’s Most Critical Technology,” which comes out Oct. 4. “Chip War” outlines the nature of the coming battle over semiconductors, showing how the power to produce leading-edge chips fell into the hands of just five companies: three from the U.S., one from Japan, and one from the Netherlands.
In an interview with Protocol, Miller touched on the evolving military uses of semiconductors, Russia’s chip procurement problems, and whether onshoring chips won’t necessarily increase the likelihood of a conflict between the U.S. and China.
This interview has been edited and condensed for clarity.
You write about how the Persian Gulf War became a pivotal moment for understanding the role microchips would play in warfare, especially with guided missiles. Are missiles still the use-case where the military needs the most advanced chips, or has that shifted to other areas like artificial intelligence or cyber attacks?
I wouldn't necessarily think about individual weapon systems. Chips are important in those weapon systems, but there's also the broader integration of sensors to decision-makers, communication technologies, [and] weapon systems — all of which rely on chips of different sorts.
So it's the integration that matters as much as the capabilities of any individual system. But as more and more systems start to become more autonomous, the demands for computing power, memory, and communication bandwidth increase on all of these systems.
So yes, semiconductors are still very important in terms of advances in specific types of munitions, but it's also [about] the ability to integrate them, to transfer information between them, and to have them working together — and all that requires advanced semiconductors at every stage of the process. Militaries are confronting semiconductors everywhere they turn, and their ability to access the right types of semiconductors and more advanced chips is crucial for every aspect of the modern battlefield.
What’s Russia’s status with the ability to procure semiconductors? There were reports that Russia was struggling to access chips for the Ukraine conflict.
Russia does have some domestic chipmaking capacity but it's very, very limited, both in terms of the level of advancement — which is not that advanced — but also in terms of capacity.
What we've learned from the war — which confirms what we presumed before the war — is that, when you take apart the Russian military system, you'll find lots of foreign-produced chips, and often American-produced chips, inside. Many of these chips are off the shelf — things that you can acquire commercially.
In some ways, the fact that every military in the world is trying to use commercial chips in their systems has been an equalizing factor. But for Russia, it has presented some challenges too. Russia never really knows whether the chips it's buying have been sabotaged in any way. Every country has that to a certain extent. But the U.S., for example, is relying on chip fabs in the U.S. or in friendly countries — whereas Russia is relying almost exclusively on chips produced in unfriendly countries.
The systems integration capabilities are also lacking in Russia. They've got some, but there's just a lot more expertise in places like the U.S. or in East Asian countries [for] how you bring different types of chips plus software plus the hardware all together in a coherent system. And I think that's also somewhere we've seen Russia face some difficulties.
Photo courtesy of Chris Miller
How successful do you think the Chips Plus Act will be?
There's a question about getting the most bang for our buck out of the Chips Act fund. I think there are ways we can get more or less lucky on that front. But I don't think there's really a serious argument that the Chips Act funding is not going to increase the amount of leading-edge chipmaking capacity in the U.S. We've already seen a number of new facilities being built on the expectations of those funds.
Does onshoring chip production increase the chances of conflict between the U.S. and China? It seems the U.S. is onshoring production in anticipation of greater conflict, but is it also possible it increases the chances of something happening by reducing the cost of conflict?
This sort of mutually assured economic destruction thesis — the idea that the U.S. and China or China and Taiwan could never find themselves in a conflict because it would be too costly to do — I think history provides many examples of times in which economically integrated countries ended up in conflict. So that's one reason to be skeptical of it.
The second reason is that the assumption that was true for most of the period from Deng Xiaoping forward in China — that Chinese leaders were primarily focused on maximizing GDP — that assumption is no longer valid. The last three years of Chinese policy have not been about maximizing GDP but about political goals and zero COVID. It doesn't mean that they're not going to care at all about economic issues. But we shouldn't assume that either China or the U.S. or Taiwan is approaching these issues solely from the perspective of making sure trading continues unharmed.
The lesson of 2022 is not that economically integrated powers can't quickly decouple, because we've seen that in a shocking way between Russia and Europe. So I don't think we should take much comfort in the fact that there's a lot of economic integration between China, Taiwan, the U.S., and other countries in the sense that is going to guarantee a peaceful outcome. I don't think history suggests that's true.