The EU aims to catch up with China and the United States in the race for the most advanced processors. Its attempt, though, shows serious blind spots.
 |
| EU Chips Act and China's Underappreciated Power |
The most recent chips, sometimes known as " cutting edge ", are frequently at the focus of current attempts to increase technical competitiveness and supply security in the field of semiconductors. The importance of older semiconductor facilities is overlooked by contemporary legislative efforts such as the European Act draft. China, on the other hand, has a stronger position in this regard.
The various chips that are incorporated into contemporary and digital life, from vehicles to computers and servers to medical equipment, are at the core of them. Because of constant technological progress, some of these semiconductors are subject to enormous pressure to innovate, which follows Moore's law: chip manufacturers try to double the circuit density and thus the computing power every two years in order to continue to produce "cutting-edge-Chips" - i.e. state-of-the-art chips. This concept applies to processors in particular, which make new smartphone or server generations more powerful. The structural width of the transistors offers information on the processor's processing capability, which is measured in "nanometers" - one millionth of a millimeter.
As a general rule, the smaller the size, the more contemporary the production method and the greater the performance. Yet, this is only one type of semiconductor. There are numerous other chips with irreplaceable functions in our modern society: chips for charging the smartphone battery, controlling the brake system, activating the airbag, reading fingerprints, or powering the server - these are not about computing power, but about a wide range of physical properties.
These chips, unlike the pure arithmetic processors discussed above, are not made in the most advanced semiconductor manufacturers. Rather, their production is dispersed throughout a plethora of smaller, generally older semiconductor plants. And the majority of these factories are located in China. When it comes to China's competitiveness in the semiconductor business, the cutting-edge chips stated before are now receiving a lot of attention. It's easy to ignore the fact that the country, along with Taiwan, is a world leader in the production of several semiconductors in older facilities. This demonstrates that judging China's technical competitiveness purely on cutting-edge chips is inadequate.
The importance of a country's role in semiconductor value chains can only be assessed in terms of specific semiconductors, their functions and end-user industries.
Chips with a wider structural width are critical for Europe's client industries in automotive, medical devices, and Industry 4.0. These are even more crucial than the previously described cutting-edge chips. Hence, in order to analyze and eliminate dependence, it is critical to understand how present manufacturing capacity are divided in order to increase Europe's resilience and secure supply of strategically crucial semiconductors. Recent legislative proposals in industrial policy, such as the EU Chips Act, do not yet incorporate this information. One example is Europe's reliance on ageing manufacturing in China and Taiwan.
It's helpful to zoom out a little to see how the fabrication of chips with wider structural widths fits into the bigger picture of semiconductor production. This displays a highly sophisticated and specialized structure with a plethora of tightly interconnected gears. A contemporary semiconductor requires about 1000 process steps, 80 distinct types of production equipment, and up to 400 different chemicals. The United States, Taiwan, South Korea, Japan, China, and Europe are all heavily involved in this intricate process, with each holding a vital, if not irreplaceable, role at various places along the value chain.
Moreover, chips differ substantially in their purpose in the end product - a current processor in a smartphone has nothing in common with a power semiconductor for charging an electric automobile. It is essentially a network of value chains that are built on closely coordinated production processes, from the materials to the machines, depending on the function and technology of the chip.
A contemporary automobile has around 1000 semiconductors, such as microcontrollers, power semiconductors, sensors, and radio transmission chips. They operate the engine, brakes, and airbags, assist with parking and navigation, receive internet, and manage the windshield wipers. Sophisticated CPUs and AI chips are becoming increasingly important for future technologies such as self-driving cars. These chips supplement, but do not replace, the necessity for chips in current vehicles. The idea that older technologies are being replaced by contemporary chips with decreasing structural widths does not apply to many of these automotive semiconductors.
This is just not always achievable because of their technological function. According to market experts, by 2030, every second semiconductor put in a car will be based on so-called mature nodes, or chips with structural widths of at least 28 nanometers. The scenario is comparable in the healthcare, engineering, and defense industries. During the 2020 shortages, older production methods for 40 to 180 nm semiconductors were especially unavailable. At times, whole production lines for automobiles, medical gadgets, or industrial machinery came to a halt.
For more than a decade, China has been pursuing the objective of developing its own, local, and competitive semiconductor ecosystem and becoming globally independent – with little success in terms of current manufacturing methods. This is mostly owing to the transnational value chain's high degree of specialization, which necessitates the development of process expertise and large investments in a single production or process step over decades. For example, the much-discussed EUV exposure machine from the Dutch firm ASML, which is crucial for cutting-edge chip fabrication, is built on more than two decades of research and development and a network of over 5000 suppliers. High subsidies alone will not cover this imbalance.
Added to that export restrictions that the USA imposed on China in October 2022 in order to slow down the country's production of state-of-the-art semiconductors or to freeze its capabilities in this area.
As a result, China is still many years behind in terms of cutting-edge semiconductors. In terms of production capacity, the country's technical advancement in the domain of mature nodes, on the other hand, is exceptional. The entrance hurdles to the market are lower than in the cutting-edge domain. This is because American export rules do not apply here, and a comparable degree of invention, money, and process understanding is not required. China has unrestricted access to the machinery, development software, and contract manufacturers it requires. Taiwan leads in production capacity in the range of 20 to 45 nanometers, followed by China, which has more than three times the manufacturing capacity of Europe. Even older technology between 50 and 180 nanometers show a substantial difference.
China's manufacturing capability in this area is more than double that of Taiwan. In such cases, Japan, Europe, and the United States cannot do without manufacturing capacity in China and Taiwan, even if all three nations operate part of these older facilities locally.
The global distribution of manufacturing capacity for mature nodes demonstrates that Europe's strength in this area has long been challenged by Chinese and Taiwanese companies, or European semiconductor companies have previously decided to relocate their production to the Asian region, for example, due to low labor and energy costs. And China, in particular, is actively working to develop this sector of manufacturing. SMIC, China's premier contract manufacturer, has announced the opening of three new mature node plants in Tianjin, Shanghai, and Beijing. Furthermore, the most recent US export constraints contribute to making older manufacturing an appealing field for China to grow its already dominant role unhindered: As a result, Europe may become increasingly reliant on mature Chinese nodes in the future.
Nevertheless, because these older production technologies do not qualify for subsidies for so-called new semiconductor manufacture, this element is presently not taken into account in the European Chips Act draft. Given the vital importance of these chips in the automotive, mechanical engineering, and healthcare industries, this gap in EU chip regulation seems inexplicable. It might be worthwhile to improve this: Meaningful discussions concerning the supply security of end-user sectors must also consider the critical importance of mature node manufacturing and draw inferences from dependence in this area. In order to strategically position itself, Europe must consider the entire spectrum of chip manufacture.
The author Julia Hess is a project manager at German think tank SNV.