EU Chips Initiative Seeks €80m to Drive Innovation in AI Computing

EU Allocates €80m to European Organizations Advancing AI Hardware and Infrastructure

The European Union has earmarked €80m in new funding to support the development of AI hardware and infrastructure in Europe. This investment aims to propel AI compute technologies from advanced chip designs through testing stages towards deployment.

The Chips Joint Undertaking (Chips JU) has introduced two linked funding opportunities that cover various stages of the AI compute value chain. One call, with a budget of €10m, will support European AI chip demonstrators. The larger €70m program will focus on establishing the necessary infrastructure for testing these technologies and procuring pilot systems.

The deadline for both funding opportunities is set for 27 October 2026 at 17:00 Brussels time. This initiative aims to foster collaboration among chip designers, AI infrastructure operators, and cloud providers in a shared development and evaluation process.

Enhancing Connectivity Between Chip Development and Real-World Testing

The approach taken by the EU involves a seamless progression that integrates chip development with infrastructure testing. Rather than treating these as separate activities, the funding opportunities are designed to bridge the gap between them.

Through the first call, advanced European AI chips will be transformed into functional demonstrators. These systems will then undergo evaluation on a common platform established through the second call.

This evaluation process will identify technologies that can transition from individual chip demonstrations to rack-level systems and, ultimately, pilot deployment.

The first call, focusing on AI chip demonstrators for EU compute infrastructure, offers an indicative EU contribution of €10m. Project proposals can request up to €1m, with flexibility in the funding amount.

Supporting Advanced European Chip Designs

The funding targets AI chip designs that have made substantial progress towards production. Applicants must collaborate with an EU-based AI chip design company responsible for the chip and its intellectual property.

Requirements include having the technology taped out at a commercial foundry, engineering samples ready or expected within the specified timeframe, and a PCB hosting the chip in production or planned during the development period.

The funding will cover the completion of the demonstrator, encompassing elements such as the chip, PCB, cooling, interconnects, software integration, and preparation for performance benchmarking.

Following completion, each demonstrator will undergo up to two months of assessment on the shared evaluation platform. Exceptional technologies may receive the AI Compute Excellence Label, a prerequisite for participation in the subsequent A2 rack development phase.

€70m Program Focuses on Infrastructure and Procurement

The second funding opportunity, with an indicative EU budget of €70m and detailed information available here, emphasizes creating an environment for testing, comparing, and deploying AI compute systems.

It will establish common evaluation infrastructure and services for benchmarking systems developed under the initiative, along with supporting the procurement of pilot AI compute racks.

Consortia applying for this opportunity must include at least one contracting authority or organization operating an EU AI Factory or similar public AI/HPC facility. Additionally, they must involve two EU-established cloud service providers representing at least three different EU Member States or EEA countries.

Advancing from Demonstrators to Multi-Rack Systems

The procurement phase aims to move beyond singular hardware demonstrations. Selected systems will be installed and optimized as scaled, multi-rack clusters, enabling the testing of distributed AI workloads across larger infrastructure.

This comprehensive action plan spans 42 months and progresses through three interconnected stages: testing A1 chip demonstrators, preparing and evaluating A2 rack systems, and procuring and installing chosen systems.

This roadmap offers a holistic approach to advancing European AI compute, starting from chip-level technology and extending through benchmarking, rack development, and infrastructure deployment.

The program is divided into two operational streams: Stream A for European AI chip and rack development and Stream B for evaluation and procurement frameworks. Currently, only the A1 demonstrator call and Topic B evaluation and deployment call are open, with the A2 rack development opportunity scheduled for a later phase beyond 2026.