Microprocessors are a strategic component and the current shortages have slowed down or stopped the production of a large number of companies. This observation has led the United States, China and the European Union to adopt industrial policies aimed at securing their supplies. In particular, the ” European Chips Act proposes to ensure the technological sovereignty of EU countries in this area. The objectives of this law seem particularly ambitious. It intends to make up for the technological delay of the European Union in the design and manufacture of microprocessors. It also aims to double the market share of European production in the global microprocessor market. The strategic objectives of the European Chips Act are therefore clearly established. However, the economic viability of this plan is questioned.
Proposed by the European Commission in February 2022, the ” European Chips Act » is justified by shortages of microprocessors due to the covid-19 pandemic. The supply of this component has suffered greatly from the occasional closures of manufacturing sites and the slowdown in logistics chains caused by the pandemic.
The European Commission intends to respond to this short-term problem with a vast structural plan aimed at drastically developing European production of state-of-the-art microprocessors. Thus, the ” European Chips Act » provides a legal basis for EU Member States to use subsidies to attract manufacturers of state-of-the-art microprocessors. A European budget of €43 billion over 5 years should subsidize the installation in the EU of manufacturers using technologies that are absent in our union. In this context, the French government last summer welcomed the alliance between STMicroelectronics and GlobalFoundries with a view to building a ” mega fab » in Isere. For its part, the Bundestag announced that it would finance up to 40% the installation project, valued at 17 billion euros, of an Intel manufacturing site in Magdeburg.
The weakness of the European budget has already been underlined. What does the €43 billion of the European plan represent when TSMC announces an annual investment budget of $30 billion? How can we catch up and ensure our technological independence with such a budget when Intel devotes $15 billion per year to its R&D?
This budgetary weakness is far from the only sticking point of the EU plan. The structure of European demand and the characteristics of the microprocessor market call into question the interest of ” European Chips Act “. We present here the main questions raised by this plan.
European industry is not demanding advanced technology, and it is not alone!
The main European manufacturers of microprocessors engrave chips in 40 nm while the world leaders in the sector have been developing the 7nm format since 2015. 2018 the end of the development of this advanced technology, preferring to focus on more promising formats. For its part, Intel intends to share the risks and relies on partnerships with founders to offer 7nm chips. Thus, from a purely technological point of view, Western companies seem to be doing less well than Samsung and TSMC. But what is the reason?
The production of computers, smartphones and more generally devices related to information and communication technologies is located in Asia. The Korean Samsung and the Taiwanese TSMC are developing the formats best able to meet the demands addressed to them. The high levels of coordination and integration required along the value chain of these products favor Asian microprocessor manufacturers. These are the drivers of technological progress serving specific reputable customers. technology takers.
The manufacture of computers and smartphones in Europe is anecdotal. Should we imagine that the installation of ” fabs at the cutting edge of technology be the basis of a future ecosystem dedicated to information and communication technologies? Where are the European companies likely to supplant Lenovo, Samsung or Foxconn?
The absence of European champions of information and communication technologies calls into question the nature of the partnerships forged with GlobaFoundries for France or Intel for Germany. Are the two American companies engaged in technological partnerships with European states? Do they seek to share the risks inherent in the development of new technologies?
Or have they agreed to set up in Europe to get closer to customers who are looking for already mature technologies? Are they looking for markets in Europe for old technologies?
The European Union must certainly secure its supplies of microprocessors. However, the needs of our industry differ greatly from the requirements of Chinese, Taiwanese or Korean companies and the search for state-of-the-art microprocessors should not be our priority.
Characteristics of the microprocessor market
Microprocessors, a cyclical market
Since the rise of microcomputing in the 1980s and throughout the development of information and communication technologies from the 1990s, the microprocessor market has been characterized by cycles of expansion and contraction in demand and prices. Indeed, the cyclical nature of demand comes up against the rigidity of production and the adjustment from one to the other takes place first of all through prices. Thus, smartphone sales once again fell to 3th quarter of 2022, falling to their lowest level since 2014. This bearish phase is expected to continue and Samsung has lowered its sales target for 2023 by 40 million units (260 million vs. 300 million). Given the installed production capacities, the prices of the chips used in the manufacture of these devices should therefore fall next year and the supply for this component is largely assured. In other words, as companies and governments around the world invest hundreds of billions of dollars in response to the current shortage, there is a real risk of overcapacity for certain types of chips. The expected returns on these investments would then be jeopardized.
Costly adaptation of production to changes in demand:
We have underlined the rigidity of the production volumes of the ” fabs “. These are also not very adaptable: modifying the characteristics of the microprocessors to be produced is a long (several months) and expensive process. Until now, the concentration of the supply of microprocessors made acceptable the investments to be made to adapt the supply to the evolution of demand. The proliferation of production sites induced by (supra) state commitments upsets this balance. By lengthening the time for return on investment, the deconcentration of the supply of microprocessors could generate a slowdown in the speed of commercialization of technological progress. The business models of fabs » but also those of producers of technological goods or distributors are therefore impacted by the new deal imposed by new entrants.
Constant R&D efforts
Upstream of manufacturing, the design of microprocessors requires colossal and constant investments in R&D. Thus, Intel devotes $15 billion each year to R&D. However, this company is currently lagging behind Samsung and TSMC in technology, which are planning annual investments of $30 billion.
The microprocessor market is therefore technologically very active. To exist on this market, and even more so to achieve the 20% market share targeted by the European plan, annual investments of several tens of billions of euros must be provisioned. The volume of this investment is much higher than the capacities of the European companies present on this market. For example, STMicroelectronics forecasts sales of $15 billion for 2022 and the partnership with GlobalFoundries should allow it to cross the $20 billion turnover mark. To remain competitive, STMicroelectronics must rely on regular technology transfers from GlobalFoundries to Europe. Thus, the French site materializing the alliance between the two companies will be managed by the personnel of GlobalFoundries who are delighted to strengthen their position as the leading foundry of semiconductors in Europe thanks to this alliance. France’s achievement of technological sovereignty in the field of microprocessors is dwindling and the interest of a financial commitment by the State in this affair is questioned.
We tried to understand the interests associated with the ” European Chips Act by comparing the announcements of the various stakeholders with a certain number of economic realities which seem objective to us. The lack of coherence between the first and the second leads us to think that political communication has considerably blurred the real objectives of the European plan. This responds to a strategy of securing supplies of microprocessors using a mature technology and heavily consumed by European industry. Far from ensuring the technological sovereignty of the EU, this plan reinforces our technological dependence vis-à-vis the United States and can contribute to the deepening of the geopolitical alignment of the European Union with this country. In a context of worsening commercial and political relations with China, this choice has very real consequences, which the Netherlands has already experienced. Indeed, Bloomberg reports that the US administration has asked ASML Holding NV to stop selling microprocessor manufacturing equipment to China. Using US patents in its production, ASML accepted the US application. In other words, the technological dependence induced by the European plan constrains export opportunities for European products.
L’ ” European Chips Act ” also generates a race for state subsidies between the different countries of the European Union, each wishing to host a ” fabulous “. The STMicroelectronics – GlobalFoundries partnership responds to the establishment of Intel in Magdeburg and Italy is negotiating the establishment of a “ mega fab in Veneto. Spain and the Czech Republic, both major car producers, certainly want to secure their supplies and we are betting on the imminent establishment of ” fabs in these two countries. Given the cyclical nature of the microprocessor market, the proliferation of manufacturing sites should increase the problems of overcapacity periodically encountered by chip producers.
Finally, centered on securing the supply of mature components, the European plan seems to forget the emergence of a disruptive technology which will upset the architecture of a large number of products over the next decade: quantum processors. We believe that the ” European Chips Act recognizes Europe’s technological backwardness and traces the path of its dependence on American technology.
Article written by:
Eric BrownAssociate Professor – INSEEC Bachelor
Pascal MontagnonDirector of the Digital, Data Science and Artificial Intelligence Research Chair – OMNES EDUCATION
<<< Also read: Attention, the agricultural crisis is not global, but European! >>>