Europe’s SiC Capacity Is Growing, but Customers Still Decide Where the Fabs Land

November 9, 2026

Image: ChatGPT

Silicon carbide is moving from a niche power material into a strategic industrial platform, but the business case is becoming more uneven. Demand for wide-bandgap power devices, combining and , is expected to exceed $14 billion by 2031, with growth of about 22% per year. Electric vehicles are shifting toward 800V architectures, renewables and storage continue to expand, and data centers are redesigning their power chains around 800V , pushing upstream toward grid-facing stages.

At the same time, price pressure is severe. Chinese producers have driven six-inch substrate prices down from around $1,500 to $500 or less. Wolfspeed, long one of the market leaders, still carries the burden of aggressive capacity expansion. After filing for Chapter 11 in June 2025 with about $6.7 billion of debt, it re-emerged with 70% less debt, but its latest quarter still showed a negative gross margin of about 13% on a non-GAAP basis. The company said underutilization remained the main drag and indicated that 25% to 30% utilization at its Mohawk Valley fab would be enough to reach gross-margin breakeven.

Europe’s position is contradictory. European companies such as ST, Infineon, and Bosch hold roughly half of the global device market, but only about 10% of the substrate market. Europe also represents only around 12% of global demand in cars, compared with about 63% for China. That mismatch is visible in investment flows: of more than $30 billion in announced investment by mid-2026, about half is tied to European companies, yet only around 40% of that money is being spent in Europe itself.

The strongest European projects are often outside the region. Infineon is building in Kulim, Malaysia, Bosch in Roseville, California, X- in Lubbock, Texas, and ST has a stake in a joint venture with Sanan in China. Chinese and Japanese companies invest mostly at home, while U.S. companies largely do the same. Vishay remains an exception after buying Newport Wafer Fab in Wales in 2024 to establish production there.

Policy has reinforced the concentration. The European Commission has approved 19 first-of-a-kind production projects for state aid under the Chips Act since 2023, worth more than €34 billion in total. Eleven are in Germany, most of the remainder in Italy, France, and Austria, with almost none in Europe’s peripheral countries. One listed exception was onsemi’s expansion in Rožnov, Czech Republic, but that project was later abandoned after changes in the market and project economics.

In , the line between and foundry matters more than in silicon. The process is closely tied to the product, and performance depends on details that are hard to separate from device design. Open capacity remains scarce. The main options named are X- in Lubbock, several Taiwanese players planning 8-inch lines, research lines such as Fraunhofer IISB, and Clas-, which recently announced a £12 million expansion. , GlobalFoundries, and Tower are not offering , and the ecosystem around open capacity is still thin.

A Bulgarian initiative to create Green Silicon Carbide Manufacturing as an integrated device manufacturer is trying to enter that gap. The project has assembled partners including Clas- for the technology license, Synopsys and SemiWise for development, and Milara for process automation. The case for such a fab depends on phased buildout, on customers beyond European automotive demand, and on leaving room for fabless innovation. In , capital can buy tools and wafers, but not the learning curve that comes from yield, reliability history, and customer qualification.