GlobalFoundries targets 7nm-class performance in Dresden
The chipmaker presents a roadmap for energy-efficient AI chips, but production is not due to begin until 2028, according to plan.
Listen to this article
There is no audio version yet. Request one and an AI voice will read the article aloud.
Read by an AI voice.
GlobalFoundries wants to develop new chip technology at its Dresden plant that should deliver the digital performance of the 7nm class. This concerns a future design target, not an already proven or commercially available 7nm production line.
GlobalFoundries presented the FDX Fusion technology on Friday. The company wants to develop it and begin production in Dresden from 2028. The technology is intended for so-called physical AI: systems that not only compute, but also combine sensors, control, communication and local processing.
According to the announcement, the 7nm-class designation refers to expected digital performance. This does not automatically mean that the chips will be made using the same production process or with the same transistor density as conventional 7nm chips from the FinFET or gate-all-around families. Tech medium Golem stresses that GlobalFoundries has not published independent measurement results from an existing FDX Fusion chip.
FDX Fusion builds on FD-SOI, a technology that uses an insulating layer beneath the silicon. Alongside digital logic, GlobalFoundries mentions analogue and mixed-signal circuits, embedded non-volatile memory and radio-frequency capabilities. This should make the technology suitable for radar, wireless communication, robotics and industrial automation, among other applications.
The roadmap provides for demonstrators for early customer testing in early 2027. An initial process design kit, which allows chip designers to prepare their designs for the production line, should become available in mid-2027. Mass production in Dresden is scheduled for 2028. All three milestones are company targets, not achieved results.
The announcement coincides with the expansion of the Dresden plant. GlobalFoundries previously announced an investment of €1.1 billion and is aiming for annual capacity of more than one million 300mm wafers by the end of 2028. The expansion is linked to the European Chips Act and German and European support schemes.
For Europe, the plan is relevant because it could bring the production of specialised chips closer to European carmakers, industrial companies and defence suppliers. At the same time, it is not yet clear which specific products customers will have made. GlobalFoundries refers to customer evaluations but has not announced any definitive production orders for FDX Fusion.
One story, several perspectives
What is established
- GlobalFoundries has announced a new FD-SOI roadmap for Dresden.
- The company cites 7nm-class digital performance as a target.
- Demonstrators are planned for 2027 and production for 2028.
- The plant is being expanded as part of European and German support for the semiconductor sector.
Left
Arguments Public support for chip production can be justified if it creates jobs, knowledge and strategic autonomy, but companies must demonstrate that the investment also delivers a social return. Subsidies should not mainly shift commercial risks onto taxpayers.
Values Public direction, good jobs, sustainability and accountability for state aid.
Consequences Conditions concerning labour rights, energy use and knowledge sharing could increase the social benefits, but may also make projects more expensive and slower.
Centre
Arguments Having a European chip base of its own is sensible for critical applications, as long as support remains targeted, temporary and subject to oversight. The government can support infrastructure and research, while companies remain responsible for commercial viability.
Values Strategic resilience, innovation and institutional oversight.
Consequences A targeted approach can reduce dependencies without fully replacing the market.
Right
Arguments The market should determine which chip technology is viable. Governments can set the framework, but should prevent subsidies from keeping projects afloat that prove technically or commercially uncompetitive.
Values Market discipline, entrepreneurship and limiting public spending.
Consequences Less subsidy reduces the risk to government, but could also leave Europe more dependent on manufacturers outside the region.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The article carefully distinguishes between an announced roadmap and proven production. The independent source supports the key caveat about the meaning of 7nm class.
- confirmed GlobalFoundries announced FDX Fusion for development and production in Dresden. — This is stated in GlobalFoundries’ official announcement. source
- confirmed The technology is intended to deliver 7nm-class digital performance. — GlobalFoundries uses this wording itself in the announcement. source
- confirmed Production is scheduled for 2028. — The official roadmap cites 2028; Golem confirms the same timetable. source
- confirmed No independent measurement results from an existing FDX Fusion chip have yet been published. — Golem describes FDX Fusion as a newly announced process and calls the performance a company description. source
- confirmed GlobalFoundries previously announced an investment of €1.1 billion in Dresden. — This is stated in the company’s earlier official announcement. source
Editor's note
The announced performance, timetable and expansion are company claims and roadmap targets. No independent verification of the eventual performance or production yet exists.Sources
More on this in Dutch media
- NOS — „globalfoundries dresden”
- Het Parool — „globalfoundries dresden”
- NRC — „globalfoundries dresden”