Technician walking in hallway of server room_PPT

11 August 20255 minute read

European Data Centres Trends

Artificial intelligence is propelling the European data centre market into a new era. Here are six key themes emerging across this landscape:

 

1. AI Creating a Data Centre Boom

AI needs far more computing power than traditional applications. AI is multiplying the demands placed on global data centre infrastructure and the growing European data centre market. Operators are having to respond to increased computing capacity, cooling systems, specialised chips and larger storage and campus requirements.

In the US, AI-specific mega campuses, such as OpenAI and Oracle's USD500 billion ‘Stargate’ campuses in Texas, have emerged to meet these demands1. The UK is building on this blueprint with its ‘AI Growth Zones’ strategy, building hubs to develop more AI infrastructure and speeding up planning proposals to accelerate this2. One of Europe's largest data centre campuses will be developed in Wales by Vantage Data Centres, set to invest over GBP12 billion in data centres across the UK.

 

2. AI Reshaping Infrastructure

The shift from air to liquid systems for cooling chips is trending. Liquid systems can more easily absorb and transfer heat, creating an efficient cooling environment primed for AI workloads.

UK electricity costs are around 10% higher than in other European countries. Operators are having to make tough decisions on location strategy and power sourcing. As a result, Small Modular Nuclear Reactors (SMRs) are gaining traction as an alternative energy solution. SMRs have one-third of the power capacity of that of traditional nuclear power reactors, can be sited on locations unsuitable for standard plants, and offer a reliable, long-term energy source3. The UK Government has chosen Rolls-Royce to develop the UK's first SMRs, signalling strong institutional backing4.

 

3. Predicting the Unpredictable

Timelines between AI development, evolving at breakneck speed, and the delivery of the infrastructure are non-complementary. On the hardware side, steps such as securing planning approvals, establishing power supply and acquiring land operate on long timelines. The technology is constantly changing which causes challenges and demand for AI remains unpredictable5. Anticipating these changes is tricky as the uptake of AI could grow or stabilise, depending on whether tangible business impact from AI is created6. As many businesses are experimenting with AI, how many scale these efforts into long-term use cases remains to be seen.

 

4. Not all Data Centres are Created Equally

Different data centres will have different functions and different needs. Some centres will be used for more traditional cloud workloads, while others will be made to support high-compute environments. Developers are having to rethink how they plan for longevity.

 

5. Continuing Strain on Power Grids

Power availability is fundamental to site selection. Around 95% of power demand was related to cloud services within availability zones. Now, as newer and smaller AI companies often lack resources to develop their own infrastructure they are opting to locate in more affordable areas outside of these established availability zones. Here, power is not as readily available. With UK electricity costs being 10% higher than those in mainland Europe, operators are facing tough decisions in balancing cost, efficiency and reliability of power.

 

6. Sustainability Efforts

Growing power and cooling demands are challenging to sustainability efforts. Throughout Europe, ESG practices are continuing to be evident where there is regulatory pressure to adopt sustainable practices. Some governments are imposing requirements to reuse the residual heat generated from centres, and financiers are conducting Power Usage Effectiveness (PUE) audits as part of the due diligence process to measure energy efficiency. The legacy of data centre infrastructure is also important. New builds can make older builds obsolete. From a sustainability perspective, strategies such as recycling facilities and retrofitting them - rather than rebuilding entirely - are being recognised as environmentally conscious and cost-effective solutions.

 

Key Takeaways

The European data centre market is entering a complex time. AI is acting as both a growth driver and a disruptor.
DLA Piper has its finger on the pulse of the data centre market in locations around the world, with leading experts in all relevant fields of technology, real estate and planning. Please get in touch if you would like to learn more and to help you navigate the development of AI regulation across the globe and check out our new, free, comparative tool AI Laws of the World. You can also explore our latest report, Navigating Global Growth in Data Centres - Riding the AI Wave, for more insights into market trends, investment opportunities, and the legal and regulatory challenges shaping the future.


1OpenAI, Oracle deepen AI data center push with 4.5 gigawatt Stargate expansion | Reuters
2Prime Minister sets out blueprint to turbocharge AI - GOV.UK
3What are Small Modular Reactors (SMRs)? | IAEA
4Rolls-Royce SMR selected to build small modular nuclear reactors - GOV.UK
5AI data center growth: Meeting the demand | McKinsey
6The cost of compute power: A $7 trillion race | McKinsey

Print