We are witnessing a new stage in the technological war being waged by two world powers, China and the United States. Since the beginning of August, Beijing has tightened its export controls on gallium and germanium, two critical raw materials used in the manufacture of batteries, certain semiconductors, and the latest generation of military radars. These measures come nine months after the US tightened controls on the most advanced range of semiconductors, those used in military artificial intelligence applications. While the news is a strong signal of Beijing’s determination to engage in a confrontation, or even an escalation with Washington on the issue of trade and technology restrictions, this struggle between the world’s two largest economies is a highly unequal one. This is the perspective presented by John Seaman, researcher at the French Institute of International Relations (IFRI) and author of “China’s Weaponization of Gallium and Germanium” (published this summer), in this interview with Futuribles.
A year ago, the US adopted a strategy that, by extending the existing framework of the Export Administration Regulations (EAR), tightened restrictions on the export of strategic technological components to China, covering a wide range of areas (chips, advanced semiconductors, supercomputers, etc.). Can you elaborate on the implications of this initial decision?
J.S.: Yes, the export controls introduced by the US on 7 October 2022 concern the most advanced ranges of semiconductors, memory chips, and the logic chips[1] used to manage data flows, particularly for artificial intelligence (AI) systems. The American authorities are most worried about the military uses of AI. The current policy is to no longer make any distinction between civilian and military uses of AI, since the boundary between the two is so fluid. This stance is particularly a response to Chinese initiatives to get civilian and military companies working together on these technologies. This development is very worrying for the Americans. The measures that Washington has put in place affect this type of semiconductor, but also the machines capable of manufacturing them, which are on the point of producing semiconductors measuring as little as seven or even three nanometres.
However, the US has a dominant position concerning some of the software used in the manufacture of semiconductors, and thanks to its dominant position upstream in the value chain it can block China’s access to very specific market segments, with the help of other countries. The US can put pressure on their European and Asian partners, more or less prohibiting them from exporting semiconductors (South Korea and Taiwan) or the machinery needed to build them (the Netherlands and Japan). As well as semiconductors, tools, and materials, this is also a question of know-how. All American citizens are prohibited from cooperating with China in this area.
But doesn’t China also have a strong position? In your article, you point out that China currently supplies more than 95% of the world’s raw gallium and 60% of refined germanium. Isn’t there a risk that Beijing could stifle the American economy by prohibiting access to these critical raw materials?
J.S.: China does have certain levers at its disposal, but they are much weaker, much less effective in the sense that interdependence is not structured in the same way in these cases. In the latest generation of semiconductors, Chinese manufacturers cannot produce anything below seven nanometres. Will they ever be able to? This is far from certain, because it is technically very difficult. The Americans are also dependent on the Taiwanese, Koreans, and Dutch, but here they have the advantage since they are dealing with their allies.
The Chinese export controls are therefore not a direct and proportional response, because gallium and germanium are not used in the logic chips or memory chips targeted by the American controls. It is more a signal of Beijing’s desire to escalate where it can.
However, the Chinese measures have their limits and also entail risks for China itself. Chinese companies are not yet capable, on a sufficient scale, of producing semiconductors combining the use of gallium nitride (GaN) and silicon carbide (SiC), which is a necessary step for creating the latest generation of power semiconductors for high-current or high-voltage applications (electric cars, renewable energies, etc.). They have not yet mastered this technology, and yet this is less challenging than producing a logic chip or memory chip of seven nanometres or less. However, Beijing is catching up, with Chinese companies investing in this sector and/or attracting investment from major foreign companies such as the German company Infineon and the Franco-Italian company STMicroelectronics, both of which are leaders in the manufacture of this type of chip.
There is a wide range of uses for power semiconductors, including in electric cars, and this is where, in my opinion, China could shoot itself in the foot if it goes too far with its gallium export restrictions. Since China is currently unable to produce, on an industrial scale, the semiconductors needed to meet the demand of its electric car manufacturers, the Chinese automotive industry will have to buy a huge amount of power semiconductors that have been produced elsewhere. But if China reduces its exports of gallium, producers of power semiconductors will not be able to supply the chips needed by Chinese carmakers to make electric cars. There will be a major blockage in the supply chain, caused by the Chinese policy. And this would happen at the very moment when Chinese manufacturers are preparing to unveil their technologies on the world market, and are trying to create a brand image for themselves. Few people in Europe or the US know the names BYD or Leapmotor, and even fewer are prepared to trust them on a technical level. These Chinese companies therefore have their work cut out to build a reputation with consumers, both in terms of their brand and their technology. If, at some point, they have to inform their new consumers of delays of a year to a year and a half because they don’t have the semiconductors they need for production due to blockages caused by the authorities in Beijing, this could seriously disrupt their entry into foreign markets.
If China doesn’t have an overwhelming advantage, isn’t there a risk that many client countries will want to develop some form of autonomy over the supply of critical raw materials, just as for rare earths?
J.S.: Yes, when you “weaponise” a dependency, you effectively create an incentive for the dependent consumer to diversify their sources of supply, which ultimately reduces the effectiveness of this lever. So, although China is currently dominant in this field, ten years from now — any sooner than this would be very optimistic — Europe may have developed a certain level of production in these materials, while also diversifying its supplies (by increasing the proportion that is produced outside China).
It is important to note that China’s dominance is not linked to any geological advantage, but is rather the result of Chinese policies that have encouraged the extraction and processing of these raw materials in China. In the case of gallium, for example, this is the result of support for heavy industries in the metallurgy sector (aluminium production in particular). A number of other factors have also contributed to the concentration of production in China: its less stringent environmental regulations, its low energy and labour costs, and Beijing’s use of a whole range of policies to stimulate foreign investment.
But that does not mean that these resources cannot be found elsewhere in the world. The Covid-19 health crisis and the war in Ukraine have already helped to focus people’s minds on the importance of diversification and resilience. These crises have catalysed a new attempt to develop a coherent industrial policy in Europe and the US, since public opinion is now more aware of the problems caused by dependence on China.
Interview by Morgan Paglia
This article has been translated from French by Sam Ferguson.
-
Memory chips are used to store programs, user data, operating systems, and much more, thereby facilitating the operation and storage of information in a computer or electronic device. Logic chips are used to carry out data processing and control operations in an electronic system. ↑
