Summary Under the proposed Cloud and AI Development Act (CADA), Grand Challenge 1 establishes a strategic framework to enhance the "environmental sustainability, performance and security of the Union's data centres." As detailed in Annex I, this challenge sets two specific, measurable targets for the EU: lowering the average Power Usage Effectiveness (PUE) to 1.15 and raising average server utilisation rates towards 50%. These targets are not standalone goals but are the practical implementation of Operational Objective 1 under Article 4(1), which mandates the development of energy-efficient technologies, advanced cooling, and grid-integrated energy management systems. This initiative aims to ensure that the EU's expanding data centre capacity is both environmentally responsible and resilient against physical and cybersecurity threats.

Detail

Grand Challenge 1, formally titled "Environmental sustainability, performance and security of the Union's data centres," serves as the first of eight strategic priorities outlined in Annex I of the CADA proposal (COM(2026) 502 final). It addresses a critical bottleneck in the EU's digital transformation: the need to deploy data centre capacity at scale while simultaneously addressing energy constraints, environmental impact, and supply chain security. The challenge is designed to move beyond voluntary sustainability reporting by establishing a clear, state-of-the-art benchmark for the Union's cloud and edge infrastructure.

The Core Performance Targets: PUE and Utilisation

The proposal defines success for Grand Challenge 1 through two specific, quantitative indicators that drive the entire initiative:

  1. Lowering Average Power Usage Effectiveness (PUE): The proposal explicitly sets a target to improve the environmental sustainability and performance of the Union's cloud and edge data centres to an average Power Usage Effectiveness (PUE) of 1.15 across the Union. PUE is a critical metric representing the ratio of total facility energy consumption to the energy delivered to the IT equipment. A PUE of 1.15 implies that for every watt of power used by the computing hardware, only 0.15 watts are consumed by overhead systems such as cooling, power distribution, and lighting. This represents a significant leap from current industry averages, which often range between 1.5 and 1.8 for traditional facilities. Achieving this target requires a fundamental shift in data centre design, moving towards advanced cooling technologies, waste heat recovery, and highly efficient power distribution architectures.

  2. Raising Average Server Utilisation Rates: The proposal aims to raise average server utilisation rates across the Union's data centres towards 50%. Historically, many data centres have operated with low utilisation rates due to over-provisioning, static resource allocation, and the inability to dynamically balance workloads. Achieving a 50% average utilisation rate necessitates the adoption of advanced dynamic management techniques. These techniques must balance multiple competing constraints, including energy costs, thermal limits, and latency requirements, to ensure that computing resources are used efficiently without compromising service quality.

Link to Operational Objective 1 (Article 4)

Grand Challenge 1 is the strategic vehicle for achieving Operational Objective 1 of the Cloud and AI Leadership Initiatives, as defined in Article 4(1) of the CADA proposal. While Grand Challenge 1 sets the "what" (the targets), Article 4(1) outlines the "how" (the specific technological and infrastructural measures the EU will support).

Under Article 4(1), the Cloud and AI Leadership Initiatives shall support the following specific actions to meet the Grand Challenge 1 targets:

  • Advancing Energy and Water Efficiency: The initiative will support the development of technologies such as innovative cooling systems, next-generation direct current (DC) data centres, waste heat utilisation solutions, and advanced energy storage systems. These are essential for driving the PUE down to the 1.15 target.
  • Promoting Quantum Integration: The proposal calls for promoting the integration of emerging quantum computing technologies for cloud and AI computing infrastructure operations, which may offer new pathways for energy efficiency.
  • Developing AI-Powered Optimisation: A key component is the development of AI-powered technologies designed to optimise server efficiency, utilisation rates, and overall computing infrastructure operations. This directly supports the 50% server utilisation target by enabling intelligent workload scheduling, dynamic thermal management, and real-time resource balancing.
  • Designing Grid-Integrated Infrastructures: The initiative aims to design and optimise cloud and edge AI infrastructures to ensure effective integration with energy grids and to increase their flexibility. This allows data centres to act as active participants in the energy market, potentially storing energy or shifting loads to times of high renewable availability, thereby supporting the sustainability goals.
  • Leveraging Diverse Energy Sources: The proposal encourages leveraging data centres as anchor clients for advanced energy management systems that harness diverse energy sources, including small modular reactors and clean hydrogen, alongside efficient energy storage solutions.
  • Deploying Test Beds and Pilot Lines: To validate these innovations, the initiative will deploy test beds and pilot lines to integrate and test technologies developed under the above points, covering energy-efficient semiconductor and quantum computing prototypes.

Security and Resilience as a Pillar of Sustainability

Grand Challenge 1 is not solely about energy efficiency; it explicitly integrates security and resilience as a core pillar. Annex I states that enhancing the security and resilience of data centres involves "integrating semiconductor technologies and quantum technologies designed and manufactured in the Union." The goal is to improve resistance to physical and cybersecurity threats, including targeted attacks.

This integration of security with sustainability reflects the broader CADA objective of reducing dependencies on non-European providers. By ensuring that the physical infrastructure underpinning EU AI sovereignty is built on EU-designed semiconductors and quantum technologies, the proposal aims to create a data centre ecosystem that is not only green and efficient but also robust against supply chain disruptions and cyber-physical attacks.

Implementation and Governance

The implementation of Grand Challenge 1 is entrusted to the Commission and Member States, as per Article 6. The proposal indicates that these initiatives will be supported by funding from Union programmes, including Horizon Europe and the Digital Europe Programme. The Commission is empowered to adopt delegated acts to amend Annex I to reflect market and technological developments, ensuring that the targets and methods remain relevant as technology evolves. This flexibility is crucial given the rapid pace of innovation in data centre technologies.

What this means for you

For CTOs, data centre architects, infrastructure providers, and SMEs evaluating the practical impact of CADA, Grand Challenge 1 signals a decisive shift from voluntary sustainability reporting to structured, incentive-driven efficiency mandates.

  • Investment in Efficiency Technologies: If you are planning new data centre builds or major retrofits, aligning with the PUE 1.15 target will likely make your projects eligible for support under the Cloud and AI Leadership Initiatives. This means prioritising advanced cooling systems (such as liquid cooling), waste heat recovery integration, and AI-driven facility management tools. Projects that fail to demonstrate a pathway to this efficiency benchmark may find themselves at a disadvantage in future funding calls.
  • Dynamic Resource Management: The 50% server utilisation target implies that static provisioning is no longer sufficient for EU-compliant infrastructure. Architects should focus on software-defined infrastructure that allows for dynamic workload balancing, server consolidation, and intelligent power capping. SMEs providing data centre management software should look into developing or enhancing AI modules that predict thermal and power loads to optimise utilisation, as these tools are explicitly cited in Article 4(1).
  • Grid Interaction Capabilities: Future-proofing your infrastructure involves integrating with the energy grid. This may mean investing in energy storage systems, backup power that can feed back into the grid, or software that allows for demand-response participation. Data centres that can flex their energy consumption based on grid conditions will be more competitive and compliant with the spirit of Grand Challenge 1.
  • Supply Chain Security: The emphasis on EU-designed semiconductors and quantum technologies suggests that procurement strategies should consider the origin and security of hardware components. Using EU-designed chips may become a criterion for funding or recognition under the sovereignty framework, as the proposal links security directly to the sustainability challenge.
  • Funding Opportunities: Projects that demonstrate progress towards these specific metrics (PUE 1.15, 50% utilisation) and incorporate the technologies listed in Article 4(1) will be prioritised for funding under the grand challenges. SMEs should monitor calls for proposals linked to the Cloud and AI Leadership Initiatives for opportunities to pilot these technologies, particularly in the areas of cooling, energy storage, and AI optimisation.

Common misconceptions

"PUE 1.15 is a mandatory legal requirement for all existing data centres."

  • Reality: The PUE 1.15 target is part of the "Grand Challenges" framework, which aims to support research, innovation, and large-scale deployment. It is not a direct regulatory mandate for every existing data centre to immediately retrofit to this standard. However, it sets the benchmark for what is considered state-of-the-art and will likely influence future permitting, certification, and funding criteria. Data centres in "acceleration zones" will face stricter sustainability requirements, but the specific PUE 1.15 figure is a strategic target for the Union's overall infrastructure, not necessarily a per-facility legal limit for legacy sites.

"Grand Challenge 1 only applies to large hyperscalers."

  • Reality: While hyperscalers have the resources to build large facilities, the initiative explicitly supports SMEs and start-ups through the Cloud and AI Leadership Initiatives. SMEs can participate by developing the technologies (e.g., efficient cooling, AI optimisation software, energy management systems) that enable these targets. The proposal encourages broad participation across the value chain, meaning that innovators in the SME sector are critical to achieving the Union's goals.

"Sustainability and security are separate concerns under CADA."

  • Reality: Annex I explicitly links environmental sustainability, performance, and security. The security aspect involves integrating EU-designed semiconductors and quantum technologies to resist threats. Therefore, a sustainable data centre under CADA is also one that is technologically sovereign and resilient against supply chain disruptions and cyber-physical attacks. You cannot achieve the full scope of Grand Challenge 1 without addressing both dimensions.

"Achieving 50% server utilisation is easy with current tools."

  • Reality: Achieving a 50% average utilisation rate across the Union requires a significant shift from over-provisioning to dynamic, AI-driven resource management. It involves overcoming complex technical challenges related to latency, thermal constraints, and workload variability. It is a complex engineering goal that drives the need for the specific AI-powered technologies mentioned in Article 4(1), rather than a simple configuration change.

Related

This is general information about a draft EU regulation, not legal advice.