Thales, Radiall, and Foxconn are developing a French semiconductor packaging venture intended to strengthen Europe’s position in outsourced assembly and test for advanced system-in-package components.
The Tessalia Technology SAS joint venture is being established at Le Barp, Nouvelle-Aquitaine, near Bordeaux. The partners have laid the foundation stone for the future facility, with production expected to begin before the end of the decade and annual output planned to exceed 50 million system-in-package components by 2033.
The project targets markets including aerospace, telecoms infrastructure, automotive, and medical electronics. These sectors share a need for dense electronic systems that meet demanding qualification, reliability, lifecycle, and traceability requirements. Advanced packaging is becoming central to that requirement because system performance increasingly depends on how devices are integrated after wafer fabrication.
System-in-package technology allows multiple dies, passives, and interconnect structures to be integrated into compact modules. The approach can reduce board space, improve signal integrity, shorten interconnect paths, and support higher functional density. It also moves more value and more process risk into the packaging and test stage.
The venture combines Foxconn’s manufacturing scale, Radiall’s interconnect expertise, and Thales’ high-reliability electronics and defence systems background. That combination is significant because advanced packaging requires process control, inspection, thermal management, electrical test, customer qualification, and long-term production assurance.
Backend capacity becomes strategic capacity
European semiconductor policy has often focused on wafer fabrication, but backend manufacturing is now becoming equally exposed. As chiplet architectures, heterogeneous integration, high-density substrates, and compact modules become more common, the ability to assemble, package, test, and qualify devices close to end markets becomes a supply advantage.
The problem is especially acute in high-reliability applications. Aerospace, defence, automotive, medical, and telecoms infrastructure customers cannot easily redesign around unavailable packaged components once systems are qualified. A package change can affect electrical behaviour, thermal performance, mechanical fit, certification evidence, and long-term service support. Backend supply is therefore part of product assurance, rather than a late-stage commodity service.
Europe has strong positions in research, power semiconductors, automotive electronics, photonics, sensors, and high-reliability systems, although advanced packaging capacity remains a practical constraint. The Tessalia project addresses that gap at a point in the value chain where sovereignty, delivery speed, and qualification discipline converge.
AI infrastructure is adding further pressure. Demand for high-performance compute has increased the commercial importance of advanced packaging because performance gains increasingly come from memory proximity, interconnect density, thermal control, and multi-die integration. Even where Tessalia’s initial markets are not commodity AI accelerators, the same industrial forces are shaping equipment, skills, materials, and customer expectations.
The planned scale is meaningful. More than 50 million system-in-package components a year would not make Europe self-sufficient in semiconductor packaging, but it would add capacity in a layer of production that has become too important to leave entirely offshore. The site’s location near established French aerospace, photonics, defence, and advanced manufacturing capability should also support the engineering ecosystem around the plant.
Manufacturers are now judging semiconductor supply assurance by more than wafer source or component availability. Packaging route, test location, process maturity, inspection capability, and the ability to maintain production support through long programme lives have become part of procurement risk.
Tessalia will be judged by execution. Building a high-volume, automated OSAT facility requires capital, trained staff, customer qualification, stable materials supply, equipment availability, and yield learning. If the venture reaches production on schedule, it will give Europe more control over an increasingly critical stage of electronics manufacturing and help close one of the less visible gaps in the semiconductor value chain.
