Intel Retreat Shifts Europe Semiconductor Reality

//php echo do_shortcode(‘[responsivevoice_button voice=”US English Male” buttontext=”Listen to Post”]’) ?>

When Intel announced in 2022 that it would build a €30 billion megafab in Magdeburg, Germany, and a $4.6 billion (~€4.2 billion) semiconductor assembly and test plant near Wroclaw, Poland, European leaders celebrated. Intel’s investments symbolized the prospect of strategic autonomy in chips, thousands of jobs, and global relevance in advanced semiconductor manufacturing—advancing progress toward the goals of the EU Chips Act.

To help ensure the success of the project, which Intel promised would “produce state-of-the-art wafers and use the most advanced technologies of the Intel Angstrom era in manufacturing,” the German government pledged about €9.9 billion in subsidies to Intel. In comparison, Poland pledged about €1.7 billion. The European Commission approved both amounts.

But by mid-2025, the dream collapsed. Intel had already suspended its plans in late 2024 amid worsening finances. In July 2025, new CEO Lip-Bu Tan confirmed the cancellation, citing insufficient customer commitments and the risks of allocating tens of billions to capacity that might never pay off.

Europe’s alternative bets

The withdrawal is disappointing for Germany and Poland, but the story does not end with Intel’s retreat. Europe still has active, and perhaps better-aligned, projects underway—chief among them the Taiwan Semiconductor Manufacturing Company (TSMC)-backed European Semiconductor Manufacturing Company (ESMC) in Dresden and GlobalFoundries’ continued expansion beyond its main European site in the same German city.

Frank Bösenberg, managing director of Silicon Saxony, acknowledged the setback with respect to Europe’s ambitions to play in the leading-edge arena but believes Intel’s fabs would not have been ready by 2030 anyway. “Their absence doesn’t affect Europe’s 20% production goal for this decade,” he told EE Times Europe in an interview.

“In the foreseeable future, we won’t have a leading-edge logic path, at least not on the continent,” Bösenberg said, adding that it’s a “glaring contradiction” that Europe is pouring money into Nvidia chips for AI supercomputers but lacks the fabs to make such chips domestically. “There’s still Intel in Ireland, but there’s no capacity at all for manufacturing leading-edge nodes on the European continent.”

To predict what will happen next requires an understanding that when the European Commission “approves” funding, it does not cut a check to Intel or any other company. Instead, it authorizes member states to grant subsidies that would normally be prohibited under EU competition law. Direct EU contributions do exist, but they are far smaller and are focused on R&D, pilot lines, and startups.

Intel’s withdrawal thus frees up national funds rather than EU-level resources. But for another project in Germany to benefit from Intel’s retreat, Berlin would have to design a new package and submit it for state aid approval by the European Commission. The Commission would then examine whether the support is proportionate, whether it addresses a genuine market failure, and whether it avoids distorting competition within the single market.

In the meantime, ESMC is moving full speed ahead. The joint venture among TSMC, Bosch, Infineon Technologies, and NXP Semiconductors is investing more than €10 billion to build a new fab in Dresden. The facility will produce wafers on TSMC’s 28-/22-nm planar CMOS and 16-/12-nm FinFET process technologies—mature, but highly relevant for Europe’s automotive and industrial sectors. According to a spokesperson at ESMC, the project will bring TSMC’s advanced manufacturing capabilities closer to European customers in the automotive and industrial markets.

The project is on schedule, the spokesperson told EE Times Europe. Earthwork was completed in spring 2025, foundation slabs were laid in the summer, and the first building levels are going up this autumn. Powered by renewable energy and supported by advanced water management, the fab reflects Europe’s push for sustainable chip manufacturing. It is expected to create about 2,000 direct jobs, plus support thousands more in the supplier base. ESMC is also investing in local talent, supporting Germany’s first English-language vocational program for semiconductor technicians.

GlobalFoundries, for its part, expects little direct impact from Intel’s decision. GlobalFoundries already operates one of Europe’s largest semiconductor facilities at its site in Dresden, but “Intel’s 1.8-nm fab would have been in a completely different ballpark,” Jens Drews, director of government affairs at GlobalFoundries EMEA, told EE Times Europe. “Currently, we focus on essential technologies from 22 nm and up, which will remain critical for automotive, industrial, and med-tech applications for many years to come.”

Drews acknowledged that Intel’s presence might have attracted more suppliers and extended the ecosystem, but he stressed that Dresden’s cluster is already unmatched in Europe. Within one zip code, it hosts three foundries—GlobalFoundries, X-Fab, and the forthcoming ESMC—as well as IDMs such as Bosch, Infineon, and Jenoptik, as well as research institutes including the Fraunhofer Center for Nanoelectronic Technologies. “There’s nothing comparable in Europe,” Drews said.

Bösenberg echoed the sentiment: “We [Silicon Saxony] are the leading cluster within Europe. All of our players are active in more mature nodes where there is a huge European market and [rising] demand. So we see ourselves in a very strong position. However, we’re missing relevance in the leading-edge logic types.”

According to Bösenberg, ESMC cannot fill Intel’s role: “The current site won’t be able to host EUV [extreme ultraviolet] equipment, which is indispensable under 7 nm. The large EUV machines simply wouldn’t fit into the currently planned structures. [ESMC] won’t go below 10 nm.”

What Europe should learn from this experience

Intel’s retreat has sparked debate over whether Europe should even pursue leading-edge fabs. Thierry Breton, the EU commissioner who championed the Chips Act, envisioned Europe as a player at the frontiers of 2-nm production. But Drews argues that innovation is not measured solely in nanometers. Europe, he said, should focus on three priorities:

• Disruptive innovation, or pursuing new architectures and energy-efficient designs that can rival leading-edge chips without requiring 2-nm fabs
• Lowering barriers to entry by providing startups and SMEs with affordable access to silicon, IP, and EDA tools
• Targeted sovereign production, or ensuring taxpayer-funded capacity is used first for domains such as aerospace, defense, healthcare, and critical infrastructure

These steps, he believes, would allow Europe to play by its own rules rather than chase TSMC and Nvidia in markets they already dominate.

Bösenberg added a more fundamental policy question: “If almost €10 billion were allocated for subsidizing a company, hopefully this was connected to a strategic objective. The question is, does this strategic objective still exist? And if so, what’s Plan B?” He pointed to Germany’s new National Semiconductor Strategy, which at the time of writing was expected in October 2025, as a possible source of answers.

Policymakers must strike a balance. Overly subsidizing one cluster risks redundancy, but spreading funds too thin risks dilution. As Drews put it, healthy competition—even among foundries on similar nodes—is possible and necessary.

Intel’s withdrawal is a reality check, not a death blow, for Europe’s semiconductor ambitions. It underscores the fragility of relying on foreign giants and the importance of aligning subsidies with actual market needs. Europe still aims to double its share of global chip production by 2030. That goal remains difficult but achievable—if policymakers learn from the Intel episode.

As Drews concluded, “Europe doesn’t need closed loops where everything made here stays here. It needs capabilities and capacities, global relevance, and the ability to build on its strengths. With ESMC, GlobalFoundries, and others, we have the foundations. What we need now are speed and focus.”

---