Dresden’s €5bn Fab Launch Shows Europe’s Chip Strategy Is About Industrial Sovereignty, Not 2nm

It is what Europe has chosen to manufacture there that truly defines this shift.

The new plant is aimed at power semiconductors and analog and mixed-signal chips - components that rarely feature in the global race for the most advanced processors but are indispensable to electric vehicles, renewable energy systems and the power infrastructure of AI data centres. Around Dresden, GlobalFoundries, the European Semiconductor Manufacturing Company and X-FAB are building out complementary capabilities.

Europe is not trying to reproduce Taiwan’s leading-edge logic industry. It is attempting something narrower, and arguably more realistic: ensuring that the industrial systems it already depends on cannot be brought to a halt by a shortage of specialised chips.

Dresden becomes the centre of the strategy

The scale of activity in Saxony makes Dresden more than a collection of individual fab projects.

GlobalFoundries began construction in March on the SPRINT expansion of its Dresden site, committing €1.1bn in capital expenditure. Germany has approved €495m in state aid, as part of a joint €623m package involving X-FAB. Once completed, SPRINT is expected to add more than 110,000 300mm wafers a year and take the site beyond 1mn wafers annually by late 2028.

Its technology mix is revealing. The expansion will support 22nm FD-SOI, BCD power technologies, radio-frequency components and automotive microcontrollers. These are precisely the kinds of chips needed by industrial and automotive manufacturers that cannot simply redesign their products around a different processor every few years.

The Dresden ecosystem is also becoming more integrated. Bosch, Infineon, NXP, Siemens, AUMOVIO and SpiNNcloud are among the partners connected to the GlobalFoundries expansion, while Exyte is responsible for general construction. The ESMC facility, a joint venture between TSMC, Bosch, Infineon and NXP, has completed its main structural construction and is moving into tool and system installation. Commercial production of 28nm and 22nm chips alongside 16nm and 12nm FinFET technologies is targeted for 2027, with capacity of up to 40,000 300mm wafers a month.

X-FAB is extending the cluster in another direction. Its Fab4Micro project in Erfurt will create about 3,500 square metres of cleanroom space for MEMS and integrated microsystems, with applications in automotive, medical and industrial equipment.

That diversity matters. Dresden is not betting on a single technology or customer base. It is assembling an industrial network around several categories of silicon that can be combined across the same regional supply chain.

France is pursuing a different route

The French model is less about creating a new semiconductor landscape from scratch.

At Crolles, the €7.5bn programme involving STMicroelectronics and GlobalFoundries is expanding existing industrial capacity. STMicroelectronics is moving its 300mm facility towards advanced 18nm STM32V8 microcontrollers, with initial deliveries beginning in the first quarter of 2026, while older 200mm production is gradually being phased out.

The contrast with Dresden is useful. Germany is concentrating large-scale physical expansion in an already dense semiconductor cluster. France is using an established site to move production towards newer processes.

Neither approach is simply a question of national industrial preference. Semiconductor manufacturing depends heavily on accumulated engineering expertise, suppliers, equipment networks and proximity to customers. Once those ecosystems reach critical mass, public money can do more than finance an individual factory: it can reinforce a regional industrial structure.

That is why Dresden's advantage is not merely its additional wafer capacity. Its strength lies in the concentration of companies and technologies around it. Crolles has a comparable strategic role for France, particularly in microcontrollers, power and analogue chips, but the two hubs represent different versions of the same policy objective.

Europe is trying to make supply security regional.

Why Europe is not chasing 2nm

The political temptation in semiconductor policy is to measure sovereignty by process node. A country that can manufacture the world's most advanced logic chips appears technologically independent; one producing older generations of silicon can look as though it is settling for second place.

That is a misleading metric for Europe's industrial economy.

A modern car needs microcontrollers, power-management chips, sensors and RF components. So do industrial automation systems, renewable-energy installations and the infrastructure required to supply electricity to AI data centres. These components may be manufactured on processes far removed from the leading edge, but their absence can stop an entire production system.

Infineon's Dresden investment illustrates the point. Power semiconductors regulate voltage, power conversion and current. In an electric vehicle they are part of the powertrain. In a solar installation they help convert and manage electricity. In an AI data centre, where computing capacity increasingly depends on enormous amounts of electricity and sophisticated power management, they are part of the infrastructure that makes the computation possible.

The strategic calculation is therefore straightforward. Europe can spend enormous sums trying to establish an independent position in sub-2nm logic, where global competition is concentrated, or it can use public capital to secure production of the components on which its existing industrial strengths depend.

The Chips Act strategy is increasingly pointing towards the second option.

Subsidies are becoming industrial risk-sharing

The economics explain why governments are involved.

Semiconductor fabs require enormous upfront investment, while the commercial return depends on long production cycles and demand that can change faster than a factory can be built. European state aid mechanisms reduce part of that risk and make projects viable that might otherwise face a less favourable investment calculation.

The GlobalFoundries expansion provides a clear example. Its €1.1bn SPRINT investment is supported by €495m in approved German aid within a wider European package. Such funding does not remove commercial risk, but it changes the distribution of that risk between companies and governments.

That is a significant evolution in European industrial policy.

The response to the chip shortages that disrupted European automotive production between 2021 and 2023 was initially framed around resilience. The emerging infrastructure suggests something more durable: supply-chain resilience is being treated as part of industrial security.

For automakers, the benefit of regional capacity is not that every chip will necessarily be produced in Europe. It is that a critical component is less likely to depend on a single distant production base when geopolitical disruption or trade restrictions intervene.

The defence effect is real, even if indirect

The same logic extends beyond civilian manufacturing.

Power electronics, RF components and MEMS technologies have applications in sensors, communications, electronic warfare and drone avionics. The Dresden and Crolles projects are commercial investments, not dedicated defence programmes, but a stronger European industrial base in these technologies creates additional resilience for systems with dual-use requirements.

That distinction matters. Semiconductor sovereignty does not mean Europe has suddenly acquired an independent supply chain for every component used by its armed forces. The projects strengthen particular layers of that chain.

And important gaps remain.

Regional wafer capacity does not by itself solve dependencies elsewhere in semiconductor manufacturing. The new facilities still exist within a wider international ecosystem involving raw wafers, lithography equipment and packaging. Expanding European fabrication therefore reduces exposure to some external shocks without eliminating dependence on foreign suppliers altogether.

The result is a more selective form of sovereignty.

Europe is building enough capacity in strategically important technologies to make its automotive, energy and industrial systems harder to disrupt, while accepting that complete self-sufficiency is neither necessary nor realistic.

For East Asian legacy foundries, that creates a less comfortable European market. Buyers that once had strong reasons to prioritise cost and global scale now have another variable to consider: where the silicon is made and how vulnerable that production is to geopolitical disruption.