Jeffrey Ding on Assessing the Implications of Huawei’s 7-nm Chip

Jeffrey Ding is an Assistant Professor of Political Science at George Washington University. Previously, he was a postdoctoral fellow at Stanford's Center for International Security and Cooperation, sponsored by Stanford's Institute for Human-Centered Artificial Intelligence. His research agenda centers on technological change and international politics. His book project investigates how past technological revolutions influenced the rise and fall of great powers, with implications for U.S.-China competition in emerging technologies like AI. Other research papers tackle how states should identify strategic technologies, assessments of national scientific and technological capabilities, and interstate cooperation on nuclear safety and security technologies. Jeff's work has been published in Foreign Affairs, Security Studies, The Washington Post, and other outlets. Jeff received his PhD in 2021 from the University of Oxford, where he studied as a Rhodes Scholar. He has also worked as a researcher for Georgetown's Center for Security and Emerging Technology and the Centre for the Governance of AI at the University of Oxford.
Anna Behuniak '26 interviewed Dr. Jeffrey Ding on October 4, 2023.
Photograph and biography courtesy of Dr. Jeffrey Ding on behalf of George Washington University.


What is the significance of China's reported success in manufacturing 7nm chips for Huawei? What exactly are the possible technological breakthroughs China has achieved in this particular case?

Previous to the 7-nanometer process that was demonstrated by Huawei, the most advanced chip that Huawei’s contracted manufacturer, SMIC, made was using the 14-nm process. This represents an impressive breakthrough, but there are still some limitations. One aspect that it bears on is the effectiveness of US sanctions on China's semiconductor industry. The key detail here is that the sanctions have still been very effective in terms of limiting China's access to EUV machines from the top companies, like the Dutch firm ASML. Huawei’s breakthrough was using a different process called DUV to make these seven nanometer scale chips. While you can still get to that very intricate level, the 7-nm level, with DUV machines, it takes a lot more work and it is not as economically efficient. Although it worked for this particular application, this might be the exception rather than the rule going forward.

In terms of mass production, will there be difficulties because China is using a different method of manufacturing? 

Yes, there will. Basically, you have a machine that can draw square grids but can't get that small. You just draw a bunch of grids and layer them on top of each other in diagonals which creates smaller and smaller nanometer process chips. You can't just make it with one iteration. You have to expose it one time, then move it a little bit, turn it around, expose it again. This increases the possibility of errors and you'll have a lot of failed chips and wasted investment. In technical terms, the yield rate of the chips is relatively low and the defect rate is relatively high. That's the difficulty with using this DUV process, rather than using EUV lithography machines, which China is still restricted from accessing.

What does this case tell us about the effectiveness of America's export controls in the area of semiconductors?

It's still too early to tell. Sanctions on EUV lithography machines have been effective because this is not a breakthrough that is founded on China producing its own EUV lithography machine. It does show, however, that the most cutting-edge processes might not be the most important going forward. This chip that Huawei is using, as well as its legacy chips or less advanced chips, could still be very useful for a wide range of economic applications. In general, the EUV lithography chokehold is still very salient. Sanctions seem to be effective. The question remains how broad is the impact of these types of advanced chips that can only be made through EUV lithography processes? I think this breakthrough shows it might not be that broad.  

Besides this case, how do you assess China's overall efforts to achieve technological self-sufficiency? What are China's trade-offs in this “forced import substitution” drive?

Overall, in the broader semiconductor sector, China has had limited success in achieving indigenous innovation. In fact, some of the strongest companies have been what Doug Fuller calls “hybrid companies.” They are oftentimes led by foreign directors and largely controlled by foreign investors, even though they are headquartered in China and employ most of their staff in China. These are companies that are neither fully indigenous or independent nor fully foreign and international. They sit in this gray zone which speaks to the broader issue of benchmarking technological self-sufficiency in a globalized world. China has struggled to assess foreign technology dependence and develop indicators of what success would look like in the first place. This has complicated their efforts to achieve indigenous innovation.

Referring to your article, “The Logic of Strategic Assets from Oil to AI,” could you explain briefly how microchips fit into the strategic logic framework?

The reason semiconductors and microchips are often the target of industrial policy and technological targeting is because they are strategic in all three forms of the logics I identified in that article. They are cumulative strategic in the sense that there are such high barriers to entry. It is very difficult for a country to just decide to build a firm like ASML to construct EUV lithography machines. Even the US does not have a firm that can do that. The private sector often struggles to come up with an optimal number of firms for the economy that can compete in these areas where there are high barriers to entry. Oftentimes, there is that externality where the government might need to step in to prop up some of these firms to compete in cumulative strategic industries. It's also infrastructure strategic in the sense that having a stronger, better semiconductor industry can produce positive spillovers across the entire economy. Since so many other sectors rely on improvements to chips, they can enhance the productivity of a wide range of economic sectors that rely on those chips: from the automobile industry, to the computing industry, to all these different internet-based industries. These are all economic spillovers that governments might want to pay attention to. Lastly, what we've been talking about, is dependency strategic. This is when countries are concerned about being cut off from their supply of certain chips. For example, consider the U.S.-China relationship, when in October 2022, the U.S. limited exports of high-end chips to China. This works because they can leverage certain nodes in the supply chain where chip production or chip design is concentrated. 

Would you say that the increase in strategic technological goods is dampening innovation globally? 

There are always trade-offs with industrial policy. If there weren't any geopolitical rivalries and we were living in an ideal world of economists where every country can specialize in what they are best at, and there is free trade, that would lead to more overall innovation and growth for everyone. But, because states have geopolitical rivalries and face threats from other states, they need to protect certain assets as strategic. That's why we see this attention towards restricting strategic technologies or trying to boost your industries to be more robust than your rivals, which deviates from this sort of idealized utopia of perfect economic competition and free trade.

In the future, do you see there being more or less geopolitical implications related to new technologies? 

We don't know the relationship between the U.S. and China, but also we don’t know it for South Korea and Japan either, for instance. There were many trade conflicts between those two countries, as well. I think geopolitics will always have a role to play, but it's uncertain whether these trends will intensify or abate in the coming decades.

In light of China's announced goals of technological self-sufficiency, massive investments in R&D and reported progress, what should be the US response?

First of all, it is a perfectly rational option not to respond. Other countries try to do this all the time. France has very aggressive goals for self-sufficiency in various strategic technologies, yet the U.S. is not trying to formulate a response to what France is doing. Obviously China, as a fiercer geopolitical rival, presents a different challenge. One thing to consider is the more China tries to go at it alone, in terms of technological innovation, the less effective China will be in terms of sustaining economic growth and maintaining access to key technologies. All of those rely on being looped into global innovation networks, and thus in part being dependent on global innovation networks. The U.S. is dependent on these networks as well and is also not technologically self-sufficient. On some level, the U.S. should not necessarily see China's efforts to gain self-sufficiency in technology as threatening. In fact, it might be counterproductive, and China might be shooting itself in the foot. One thing the U.S. is doing is closely evaluating its own supply chains and trying to figure out where there might be too much dependency on a single source-supplier or country. I might try to diversify its supply chains to be more resilient to disruptions. The U.S. is acknowledging that it needs to compete in a world of globalized and innovative networks. While the U.S. has to accept some level of dependency, it also needs to diversify its dependence away from a single country and have it come from many different sources. This way, no one country can have undue influence by cutting off supply in critical technologies. 

Anna Behuniak '26Student Journalist

AAAndrey A, Public domain, via Wikimedia Commons

Share this:

Leave a Reply

Your email address will not be published. Required fields are marked *