Summary Under the proposed Cloud and AI Development Act (CADA), the European Union has set a strategic benchmark to improve the environmental sustainability of its cloud and edge data centres to an average Power Usage Effectiveness (PUE) of 1.15 across the Union. This target is not a mandatory regulatory threshold for individual operators but a key performance indicator for "Grand Challenge 1" within the Cloud and AI Leadership Initiatives, as outlined in Annex I. It is paired with a complementary objective to raise average server utilisation rates towards 50%, reflecting a dual focus on cooling efficiency and computational density. Achieving these goals would require a significant shift in current industry standards, leveraging advanced cooling, waste heat recovery, and AI-powered workload management.

Detail

The target PUE of 1.15 represents a highly ambitious strategic goal for the EU's digital infrastructure, embedded within the Cloud and AI Leadership Initiatives. These initiatives are established under Article 6 of the Regulation to support research, development, and deployment activities addressing major technological and industrial challenges. The specific target appears in Annex I, under Grand Challenge 1: Environmental sustainability, performance and security of the Union's data centres.

The Strategic Benchmark: Annex I(1)(1)

The proposal explicitly defines the scope of this challenge in Annex I(1)(1), which states:

"Lowering average Power Usage Effectiveness: improving 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."

This text clarifies that the objective is to achieve this average across the entire Union, rather than mandating that every single facility must immediately meet this figure. The target serves as a north star for the initiatives funded and supported under the Act, guiding the development of next-generation technologies. The focus areas listed to achieve this include:

  • Advanced data centre energy efficiency technologies such as cooling and waste heat recovery.
  • Quantum computing technologies for cloud and compute infrastructure operations.
  • Grid integration and advanced energy management systems.
  • Pilot lines for the validation of next-generation energy-efficient technologies at operational scale.

The Dual Pillar: Server Utilisation

The PUE target is inextricably linked to a second, equally critical metric: server utilisation. Historically, data centres have suffered from low utilisation rates (often estimated between 10% and 15%), meaning vast amounts of energy are consumed to power idle servers. CADA addresses this directly in Annex I(1)(2):

"Raising average server utilisation rates of data centres: raising average server utilisation rates across the Union's data centres towards 50%, by integrating for example, AI-powered technologies for dynamic server utilisation management, runtime workload management and scheduling or for balancing utilisation, energy cost, thermal constraints, and latency requirements."

This pairing is crucial. A PUE of 1.15 is extremely aggressive by current industry standards, where 1.2 to 1.3 is often considered excellent for new, state-of-the-art builds. Achieving a Union-wide average of 1.15 requires not only superior cooling infrastructure but also a fundamental shift in how compute resources are managed. By raising utilisation to 50%, the energy overhead of cooling and power distribution is spread across more active workloads, making the PUE target mathematically and operationally feasible.

Mechanisms to Drive Progress

While the 1.15 PUE is a strategic target for the Leadership Initiatives, CADA establishes several regulatory and operational mechanisms to create an environment conducive to reaching it:

  1. Cloud and AI Leadership Initiatives: Under Article 6, these initiatives are the primary vehicle for funding and supporting the R&D required to reach these targets. They will focus on "grand challenges" that bridge the gap between current capabilities and the state-of-the-art required for a 1.15 PUE.
  2. Data Centre Acceleration Zones: Article 10 requires Member States to designate data centre acceleration zones. While the 1.15 PUE is not a direct condition for designation, Article 11 mandates that sustainability requirements for data centres in these zones must use the key performance indicators (KPIs) defined in Commission Delegated Regulation (EU) 2024/1364. This ensures that new capacity deployed in these zones adheres to rigorous, harmonised sustainability standards that align with the broader goals of Grand Challenge 1.
  3. Strategic Project Designation: Under Article 14, the Commission may designate data centre projects as "strategic projects" if they include "highly sustainable or innovative features." Projects that demonstrate a pathway to the 1.15 PUE target or the 50% utilisation rate could be prioritised for support, including streamlined permitting and access to Union funding.
  4. Monitoring and Reporting: Article 15 empowers the Commission to monitor the Union's compute capacity and the size of the capacity gap. This monitoring framework is expected to track progress against the indicators set out in Annex I, providing transparency on whether the Union is moving towards the 1.15 PUE and 50% utilisation goals.

Technological Enablers

To bridge the gap between current averages and the 1.15 target, CADA highlights specific technological avenues in Annex I:

  • Advanced Cooling: Development of next-generation cooling technologies, including liquid cooling and direct-to-chip solutions, to minimise energy loss.
  • Waste Heat Recovery: Systems designed to capture and reuse heat generated by servers for district heating or industrial processes, turning a waste product into a resource.
  • Grid Integration: Technologies that allow data centres to interact dynamically with the energy grid, reducing load during peak times or optimising the use of renewable energy sources.
  • AI-Powered Efficiency: As explicitly noted in Annex I(1)(2), the use of AI for dynamic server utilisation management, runtime workload scheduling, and balancing thermal constraints is a key enabler for reaching the 50% utilisation target, which in turn supports the PUE goal.

What this means for you

For technology leaders, data centre operators, and investors, the 1.15 PUE target signals a definitive shift in EU policy and investment priorities. While it is not an immediate legal mandate for existing facilities, it shapes the competitive landscape and defines the criteria for future support.

  • Investment Strategy: If you are planning new data centre capacity in the EU, aligning your design with the principles of Grand Challenge 1 is essential. Technologies that contribute to lower PUE and higher server utilisation will be favoured in public procurement and funding calls.
  • Funding Eligibility: Projects that demonstrate a clear roadmap to the 1.15 PUE target and 50% server utilisation may be eligible for support under the Cloud and AI Leadership Initiatives. Stakeholders should monitor calls for proposals related to Grand Challenge 1.
  • Strategic Project Applications: If your data centre project includes innovative sustainability features, consider applying for designation as a strategic project under Article 14. This can unlock streamlined permitting processes and potential financial support.
  • Compliance Readiness: Although the 1.15 PUE is a strategic target, the sustainability requirements for data centres in acceleration zones will be based on the KPIs in Delegated Regulation (EU) 2024/1364. Operators should begin measuring and optimising their PUE and server utilisation now to ensure readiness for future regulatory tightening.
  • Technology Adoption: Invest in AI-driven management tools for dynamic server utilisation and cooling. These technologies are explicitly mentioned in CADA as critical enablers for achieving the dual targets of PUE and server utilisation.

Common misconceptions

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

  • Fact: It is a strategic target for the Cloud and AI Leadership Initiatives, not a direct regulatory mandate for individual operators. However, it influences the criteria for strategic projects and the sustainability requirements applied in acceleration zones.

"Only cooling technology matters for achieving a 1.15 PUE."

  • Fact: CADA explicitly links PUE reduction with server utilisation. Improving cooling alone is insufficient; you must also increase the efficiency of compute usage to reach this target. Low utilisation dilutes the efficiency gains from advanced cooling.

"This target only applies to large hyperscalers."

  • Fact: While large providers will play a significant role, the target is for the "Union average." This means smaller, highly efficient data centres can contribute significantly to the overall average. SMEs that innovate in this space can benefit from support and recognition.

"The 50% server utilisation target is a separate, unrelated goal."

  • Fact: It is intrinsically linked to the PUE target. Low server utilisation undermines PUE improvements. CADA recognises this by addressing both in the same Grand Challenge, acknowledging that efficiency is a function of both infrastructure and workload management.

Related

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