📊 Full opportunity report: EuroHPC. The compute substrate. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.

EuroHPC’s compute infrastructure currently supports European AI projects at the mid-sized model training level, but it is not yet sufficient for frontier-scale AI training, according to recent assessments. This operational capacity is critical as Europe advances its AI independence and competitiveness.

The EuroHPC Joint Undertaking has established a network of 19 AI Factories and flagship systems across Europe, including JUPITER, LUMI, and Leonardo, which are operational and support various AI research projects. These systems enable the training of models up to approximately 70 billion parameters, exemplified by Apertus 70B on Alps.

However, the infrastructure faces structural limits in supporting the training of trillion-parameter models required for frontier AI. The €20 billion InvestAI Facility aims to build up to five AI Gigafactories with over 100,000 high-performance processors to fill this gap, with the first selection process ongoing through 2026. The June 2026 timeline for Gigafactory selection and the August 2026 EU AI Act enforcement deadline are key milestones shaping strategic infrastructure deployment.

Additionally, the infrastructure exhibits heterogeneity issues, with hardware fragmentation across CUDA, ROCm, and multiple generations of hardware, increasing software complexity and operational overhead for European AI developers. Geographical concentration of flagship systems in wealthier member states further raises concerns about structural inequality.

Implications of EuroHPC Infrastructure for Europe’s AI Leadership

The current EuroHPC infrastructure confirms Europe’s capability to support mid-sized AI models but reveals significant structural limitations for frontier AI training. The €20 billion Gigafactory plan is a strategic response to these gaps, crucial for Europe’s goal to develop independent, large-scale AI capabilities. Addressing heterogeneity and geographical disparities will be vital to ensuring equitable and scalable AI development across Europe, impacting Europe’s position in global AI competition and innovation.

EuroHPC’s Role in European Supercomputing and AI Development

Since its creation in 2018, EuroHPC JU has coordinated Europe’s supercomputing efforts, with a €10 billion budget for 2021-2027 encompassing infrastructure and AI Factories. The program’s architecture includes regional AI Factories, national gateways, and flagship systems, supporting a wide array of AI research and application projects.

Recent developments include the deployment of JUPITER (ranked #4 globally), LUMI (#9), and Leonardo (#10) in the TOP500 supercomputers, confirming Europe’s competitive position in high-performance computing. The first release of the EuroHPC Federation Platform in April 2026 marks a milestone in operationalizing this infrastructure for AI and quantum computing. The ongoing selection of AI Gigafactories aims to scale AI training capacity to frontier levels, addressing the limitations identified in prior assessments.

Despite these advances, the infrastructure’s heterogeneity and concentration in wealthier countries present ongoing challenges, which the new framework and procurement efforts seek to address.

“The EuroHPC infrastructure is operationally credible for mid-sized models but faces structural limits for frontier AI training, which the €20 billion Gigafactory plan aims to address.”

— Thorsten Meyer

Unresolved Challenges in Infrastructure Scalability and Equity

It remains unclear how quickly and effectively the upcoming AI Gigafactories will scale to meet frontier AI training demands. The impact of hardware heterogeneity and geographical concentration on operational efficiency and equitable development across member states is still being evaluated. Procurement decisions and deployment timelines may shift as negotiations and technical assessments continue through summer 2026.

Upcoming Milestones in Gigafactory Selection and Deployment

The European Commission and EuroHPC are expected to finalize the selection of up to five AI Gigafactories by June 2026. Deployment and operationalization of these facilities will be critical to meeting the EU’s AI development goals, with the first systems anticipated to become operational in late 2026 or early 2027. The enforcement of the EU AI Act in August 2026 will also influence regulatory and strategic priorities for infrastructure development.

Key Questions

What is the current capacity of EuroHPC systems for AI training?

EuroHPC systems like JUPITER, LUMI, and Leonardo support models up to around 70 billion parameters, suitable for mid-sized AI projects but not for frontier-scale models.

What are the main challenges facing Europe’s AI infrastructure?

Key challenges include hardware heterogeneity, software complexity, and geographical concentration, which may hinder scalable and equitable AI development.

How will the €20 billion AI Gigafactory plan address current limitations?

The plan aims to create up to five large-scale facilities with over 100,000 processors to support trillion-parameter models, closing the capacity gap for frontier AI training.

When will the new AI Gigafactories become operational?

Selection is ongoing through 2026, with deployment expected to begin late 2026 or early 2027, subject to procurement outcomes and technical readiness.

What impact will this infrastructure have on Europe’s AI independence?

If successfully deployed, it will significantly enhance Europe’s capacity to develop and train large-scale AI models independently, reducing reliance on external cloud providers.

Source: ThorstenMeyerAI.com