13APR

Crolles fab suspended as GF pulls back

3 min read

17:09UTC

The second of Europe's three Chips Act flagship fabs has stalled, freezing €2.9bn in French state aid.

← Back to Europe's chip ambitions meet reality Jump to analysis ↓

TechnologyDeveloping

Key takeaway

The EU Chips Act's milestone-gated subsidy design freezes aid when private partners retreat.

GlobalFoundries suspended its participation in the joint €7.5bn fab with STMicroelectronics in Crolles, France, in mid-2025. The company stated it would align expansion "with customer demand and market conditions" ¹. The pullback froze €2.9bn in French state aid, because EU rules require private co-investment milestones to be met before subsidies can flow.

The technology choice contributed to the withdrawal. The Crolles facility was planned around FD-SOI (fully depleted silicon-on-insulator), a specialised chip architecture used in automotive sensors and IoT devices. FD-SOI has a narrower customer base than mainstream FinFET (the transistor design in most modern processors). STMicroelectronics is the world's leading FD-SOI manufacturer, but GlobalFoundries' recalculation reflected the reality that demand projections for this niche node were insufficient to justify the co-investment.

EE Times identified the broader structural problem: the Chips Act's subsidy architecture creates a deadlock when private partners pull back ². Subsidies are milestone-gated, meaning the public money cannot be released to keep a project alive when the private partner withdraws. The mechanism works well in an upcycle. In a demand downturn, it freezes rather than stabilises. France now holds €2.9bn in committed state aid with no clear path to disbursement, and STMicroelectronics is left without a construction partner for a facility it cannot build alone.

Deep Analysis

In plain English

Near Paris, in a town called Crolles, France and STMicroelectronics (a major European chipmaker) planned to build one of Europe's most advanced semiconductor factories. The idea was to build it jointly with GlobalFoundries, a large American chipmaker, using a technology called FD-SOI that is particularly efficient for chips used in cars, internet-connected devices, and wireless communications. GlobalFoundries pulled out in mid-2025. The American company said demand from customers was not sufficient to justify the investment. France had promised nearly €3 billion in state aid to make the project happen; money that is now frozen because there is no factory to build. The setback matters because France had positioned this factory as central to its national plan to rebuild domestic chip manufacturing. Without a replacement partner, France will continue to depend on factories overseas for the kind of specialised chips its automotive and aerospace industries need.

Deep Analysis

Root Causes

FD-SOI (Fully Depleted Silicon-On-Insulator) is a specialised process technology. It delivers measurably better power efficiency than conventional bulk CMOS at comparable nodes, but its ecosystem of compatible IP blocks and design tools is significantly smaller than mainstream FinFET.

The customer base consists primarily of STMicroelectronics' own products, certain NXP automotive lines, and selected Samsung designs. That is insufficient volume to justify a new €7.5bn fab without extraordinary state support or a new anchor customer.

GlobalFoundries' financial position compounds the technology constraint. The company went public in 2021 at a valuation that subsequently fell by over 50%. Its capacity investments since 2022 have been concentrated in the US, incentivised by the US CHIPS Act's investment tax credit and defence customer demand.

The European project, by contrast, offered milestone-contingent French state aid with no equivalent tax-credit mechanism, meaning GF bore construction risk without a proportionate risk-transfer instrument.

A third cause is the EU Chips Act's milestone disbursement architecture. The €2.9bn French state aid was unlockable only upon verified construction progress, meaning GF needed to commit equity capital before any public subsidy arrived. For a company with constrained balance sheet capacity, that sequencing was prohibitive.

What could happen next?

Consequence
France's €2.9bn state aid allocation is stranded without a partner to receive it, requiring either a new JV partner search or a redeployment decision with new EU state aid notification.
Short term · 0.85
Risk
STMicroelectronics loses its most credible path to domestic European volume FD-SOI capacity, increasing dependence on GF's US fabs for its automotive and IoT product lines.
Medium term · 0.8
Precedent
The second major EU Chips Act flagship suspension in 12 months signals to future applicants that milestone-contingent EU state aid carries execution risk that US CHIPS Act tax credits do not, weakening Europe's position in future fab attraction negotiations.
Long term · 0.75

Source Landscape

This story draws on neutral-leaning sources

Primary parallel: The Crolles suspension mirrors the 2015 collapse of the ALTIS Semiconductor joint venture between IBM and Infineon at the Essonnes facility (also near Paris), where IBM's exit from chip manufacturing left a European partner holding sunk costs and no volume customer. In both cases, a US partner's strategic pivot; IBM exiting IDM, GF exiting European expansion; stranded a French government investment in a technology without a domestic anchor customer.

Counter-parallel: The STMicroelectronics Crolles 300mm upgrade in 2004, also state-aided, succeeded precisely because ST was the primary FD-SOI customer: the fab served its own internal demand and could therefore underwrite utilisation. The 2024 joint venture failed in part because it tried to replicate that model with GF as operator and ST as customer, but GF lacked ST's captive volume certainty.

The suspension freezes €2.9bn of French state aid that was already allocated and publicly committed. France now carries a politically visible stranded commitment with no clear redeployment target. STMicroelectronics loses a domestic volume partner and faces the prospect of commissioning future FD-SOI capacity from GlobalFoundries' existing US fabs; meaning European state aid effectively finances American supply chain resilience.

Structural impact: France's ambitions under the France 2030 semiconductor programme took their most significant single blow. The programme had earmarked ST/GF Crolles as its flagship advanced manufacturing site. Without a replacement partner or an ST decision to build independently, French leading-edge analogue/RF capacity will remain dependent on non-European foundries through the 2030s.

Consensus view: TrendForce and the European Semiconductor Industry Association both identify the Crolles suspension as a structural consequence of FD-SOI's limited addressable market: the process technology excels in IoT, automotive radar, and ultra-low-power RF applications, but its customer base is too narrow to support a €7.5bn greenfield build.

GlobalFoundries' 2024 strategic review concluded that European expansion offered lower returns than capacity additions at its Malta, New York facility, which serves a larger and more accessible US defence and automotive customer base.

Counter-view: The Institut Montaigne argues the customer demand framing is a rationalisation: GlobalFoundries' real constraint is balance sheet, not market. After a disappointing IPO and sustained share price weakness, GF lacked the equity headroom to absorb €3.75bn of construction risk on a non-US project. The demand was sufficient; the capitalisation was not.

Key tension: Whether the FD-SOI technology choice was a fundamental strategic error by the Commission's subsidy architects, or whether the failure is purely a GF capital markets problem that a differently structured partnership could have avoided.

Sources:EE Times

Mentions:Crolles →STMicroelectronics →GlobalFoundries →EU AI Act →France →Malta →Prima →Time →Infineon →FD-SOI →FinFET →

First Reported In

Update #1 · Europe's chip ambitions meet reality

EE Times· 13 Apr 2026

Read original →

Different Perspectives

European Commission

Brussels imposed €820m in platform fines, opened DMA cloud probes against Amazon and Microsoft, and issued its first Chips Act fab designations in a single period, signalling that enforcement will carry the programme the Commission's industrial investment arm could not. The quiet omission of the 20% semiconductor target from official communications is a retreat without announcement.

France

Paris is deploying Mistral as a policy instrument: €2bn+ in direct and indirect state backing, a military framework agreement requiring French-infrastructure-only deployment, and Bpifrance leading the $830m compute debt raise. The approach mirrors France's 1993 cultural exception doctrine applied to AI; defining a category of national strategic activity where market logic cannot override sovereign control.

Germany

Berlin's semiconductor strategy took its largest single blow with Intel's Magdeburg cancellation, leaving ESMC Dresden as the only proceeding flagship. Germany is compensating in AI through conditions on the Cohere/Aleph Alpha merger and Schwarz Group's consolidation of Aleph Alpha's shareholding, but the conditions risk fragmenting the combined entity's engineering operations while trying to anchor it in German infrastructure.

United Kingdom

Britain launched a £500m Sovereign AI Unit outside EU frameworks, chaired by a Balderton Capital partner, with no published investee criteria. The investment sits well below France's €2bn+ commitment; the lighter regulatory environment is the UK's real differentiator, but risks making it a gateway for US AI labs rather than a sovereign actor.

United States (USTR)

Washington filed a Section 301 investigation naming DMA cloud rules as economic warfare, treating European cloud platform regulation as a trade dispute. The probe targets cloud interoperability specifically, not the app store fines, revealing which enforcement actions Washington considers a genuine commercial threat.

European cloud industry (OVHcloud, Hetzner, Scaleway)

European cloud providers deliver 4-14 times the compute value per euro versus AWS but hold only 15% of the European market against 70% for US hyperscalers. DMA cloud interoperability mandates are the catalyst they cannot create themselves; the barrier is enterprise inertia, not price.