Following Nancy Pelosi’s visit to Taiwan, international headlines focused on the US House Speaker’s talks with President Tsai Ing-wen, as well as China’s subsequent military drills. Though widely unnoticed, she also held talks with Mark Liu, chairman of the world’s largest contract semiconductor manufacturer: The Taiwanese Semiconductor Manufacturing Company (TSMC). In a similar vein, the EU has put chip production on top of its agenda to obtain “technological sovereignty” or at least to reclaim some of the semiconductor capabilities it has lost, primarily over the last two decades. Why have EU companies failed to compete? Where are the deficiencies that ought to be addressed? How should we evaluate Brussels’ response to these issues? We shall see below.

foundry model

Why the EU lacks cutting-edge production for semiconductors

Before 2000, things were not looking that bad for the European Union in terms of semiconductor manufacturing. After all, the EU could claim more than 20 percent of global production and a similar amount of leading-edge output.

Although the foundry model certainly did contribute to the  subsequent demise, with its share of global production falling under just 10 percent by 2020, this is only part of the picture. Semiconductors require a sophisticated ecosystem, and a significant part of that gradually vanished as EU consumer products companies, such as Nokia or Siemens Mobile, lost market shares to their competitors.

As these players were increasingly replaced by American or Asian counterparts, automobile as well as industrial manufacturers turned into the major customers for semiconductor suppliers. It should be noted that the consumer electronics industry uses logic semiconductors and these profit from shrinking node sizes in the production process, thus enhancing energy efficiency as well as overall capability. This industry also generates the relevant sales volume to maneuver the enormous costs of chip design and production:

A foundry capable of producing at the 5nm node may cost around USD 20 billion with additional operating costs of around USD 1 billion per year. Additionally, to maintain competitiveness, firms must constantly invest in new equipment. The Union’s businesses have clearly fallen behind other regions, as the graph showcases.

graph: semiconductor manufacturing equipment

The reason that in 2020 almost 50 percent of the European wafer capacity consisted of 180 nm nodes or larger is that automakers and industrial applications have primarily used analog and mixed-signal semiconductors in their respective production.

Importantly, these do not profit from shrinking node sizes the way logic semiconductors do. The lack of investment is therefore not necessarily an economic miscalculation but rather the process of adapting to new market realities.

Today, the EU is not home to any leading-edge production facility. Intel’s plant in Leixlip (Ireland) and STMicroelectronics in Crolles (France) are the only facilities capable of producing the 14nm node.

Even more notable is the semiconductor value chain’s extreme concentration at 5nm node or below. Currently, only Taiwan’s TSMC and South Korea’s Samsung produce them. For the 10 nm node processors, only two more countries, the USA and Israel, join the Asian powerhouses as producers of advanced microchips.

graph: global semiconductor manufacturing capacity by location

This dependency on foreign suppliers intensifies with shrinking node sizes and thus leaves EU companies vulnerable to geopolitical tensions and external supply shocks. Likewise, the recent chip shortage has caused significant damage to the Eurozone’s economy, with an estimated loss in value creation of about EUR 11 billion.

Does the EU need cutting-edge chip production?

With the European Commission’s 2021 announcement of the Digital Compass, Brussels left no doubt about it: Not only should the EU increase its share in global chip production to 20 percent by 2030, but it should also strive to host manufacturing capacity below the 5nm node.

In the following year, the Commission’s President von der Leyen accelerated the EU’s push for “technological sovereignty” as she announced more than EUR 43 billion in public and private investment in the context of the European Chips Act.

A look into current market trends may reveal some insight into the increasing political awareness of the Union’s weak positioning. Between 2020 and 2030, the European semiconductor market is expected to grow from EUR 44 billion to EUR 80 billion.

Increasing digitalization in all areas of human life – with artificial intelligence, high-performance computing, autonomous driving, or wireless communication at its heart – will greatly increase aggregate demand for leading-edge semiconductors from around 19 percent to 43 percent by 2030.

A reasonable approach?

To assess whether the previously outlined initiatives suffice to restore “technological leadership,” it pays to take a look at the Union’s chip design industry. On the one hand, an EU leading-edge plant may not find sufficient demand to make the enormous investments necessary an economically viable decision.

While this point may appear contradictory to the outlined market developments, we need to remember that fabs (i.e. semi-conductor fabrication plant) produce for chip designers, except for IDMs, who obviously produce for themselves. Unfortunately, the EU’s overall downward spiral is also reflected in the marginal role its chip designers play as customers of advanced wafer fabrication:

Last year, the two American companies, Apple and Qualcomm, were estimated to account for about 77 percent of global  5nm shipping from pure-play foundries, while not a single Member State’s company was relevant enough to even be listed.

On the other hand, with the CHIPS and Science Act, the US has undergone great efforts to reshore production. In this context, TSMC announced in 2020 that it will build a new 5nm semiconductor fabrication plant in the US, followed by Samsung. Therefore, American fabless commercial interests in advanced manufacturing within the Union’s territory may be limited.

Policy options for the EU

For the EU to get its chips together, it should consider the complex interplay of cutting-edge foundries with chip designers, electronic design automation, chemical or equipment vendors, as well as research and technology organizations, to be found in places such as Taiwan.

Reclaiming “technological sovereignty” not only requires increasing production capacity but also re-establishing competitiveness in the wider supply chain of semiconductors, especially with respect to its design.

In concrete terms, there are three conceivable scenarios:

  1. An EU cutting-edge foundry with its own process node
  2. An EU cutting-edge foundry with licensed nodes from others, such as TSMC or Samsung
  3. An EU cutting-edge foundry operated by foreign companies

The first option is particularly hazardous because chip designers build their products on company-specific process nodes, meaning fabless companies cannot simply change between TSMC or Samsung and the potential EU fab.

The European Union is well advised to consider the financial risks associated with this approach. However, given Intel’s recent announcement of a EUR 33 billion investment in the EU microchip ecosystem, including cutting-edge production capabilities, it seems like political stakeholders are aware of this issue and tend to favor scenario three.

Additionally, cooperation with the US, whose production model is more likely to succeed given its strong business case through dominant chip designers, might help to mitigate risks associated with extreme concentration of advanced semiconductor production.

About the author

Leon Mosbacher was an intern at the Bertelsmann Stiftung, involved in the China-Europe as well as transatlantic part of the Stiftung’s Europe’s Future program. He earned his B.A. in Economics at the University of Aberdeen in Scotland; his interests include international trade and politics with a focus on Chinese economic and foreign policy. Having interned in China and with intermediate Mandarin skills, he wants to pursue research on Chinese global politics.

Read more on the EU and Asia:

A Gloomy Outlook for EU-China Relations: Divergence, Distancing, Decoupling

The Crisis in the Taiwan Strait: Another Slap in the Face for Globalization

Caught between Russia and the West? China’s Struggle for a Position on Ukraine