How Google Cloud is Accelerating its UK Data Centre Strategy

Google Cloud’s Waltham Cross data centre will become fully operational by the end of 2025.

Located in Hertfordshire, UK, the facility is set to deliver high-performance, low-latency cloud infrastructure to British businesses and represents a major milestone in Google Cloud’s strategy to support enterprise AI workloads across the UK market.

This investment forms part of Google Cloud’s significant infrastructure expansion that has reached 42 regions globally, with recent additions in Sweden, Mexico and South Africa. The company has also extended its two million mile terrestrial and subsea fibre network with new cables to support AI workload bandwidth requirements.

“Delivering on our promise, this facility will provide British businesses with the high-performance, low-latency cloud they need to compete globally,” says Maureen Costello, Vice President UK, Ireland, Sub-Saharan Africa at Google Cloud, via a Google Cloud blog post.

“We're close to reaching a major milestone with our data centre in Waltham Cross, which will be fully operational by the end of the year.”

AI workloads drive demand for specialised data centre architecture

The company delivered more than 3,000 product advancements across its cloud platform in 2024, with data centre infrastructure serving as the foundation for AI model deployment.

Notably, Google Cloud’s AI Hypercomputer architecture addresses enterprise AI requirements through custom silicon and optimised software stacks deployed in purpose-built data centres. 

The platform incorporates TPU v7 processors that provide 30 times more performance than previous generations, requiring specialised cooling and power infrastructure.

Enterprise AI applications create different demands on data centre infrastructure compared to traditional workloads. These applications require significant computational resources, high-bandwidth networking, and low-latency connectivity, particularly for real-time inference and model training operations.

The shift towards AI-first architectures is changing data centre planning. Rather than provisioning resources for peak traditional workloads, data centre operators must account for the variable computational demands and power requirements of AI training and inference operations.

Google Cloud reports more than seven million developers using Vertex AI and AI Studio, representing a 40-fold increase in Gemini usage. This growth pattern demonstrates the scale of demand driving data centre infrastructure investment for AI workloads.

Maureen highlights the infrastructure challenge at Google at Google Cloud Summit London 2025: “We’re seeing more than seven million developers using Vertex AI and AI Studio: a 40 times increase in Gemini use on the platform last year, driven by stronger adoption of Gemini Flash, Gemini 2.5, Imagen, and most recently Veo, our advanced video generation model.

“AI is assisting Google Workspace business users more than two billion times a month.”

Rising demand for data sovereignty 

Regulatory requirements are driving increased demand for data centre infrastructure that can process sensitive information within specific geographical boundaries. 

Google Cloud’s data residency commitment now includes Gemini 2.5 Flash processing, enabling organisations to keep AI model inference entirely within UK data centres.

The company offers three data centre deployment models: public cloud deployment with location controls, sovereign solutions operated by local partners and completely isolated deployments for highly regulated industries. Each model requires different data centre architectures and security implementations.

In France for instance, Google Cloud operates Trusted Cloud by Thales that meets SecNumCloud requirements, with similar deployments planned for Germany. These sovereign data centre solutions require local partnerships and specialised security infrastructure.

The UK facility too represents Google Cloud’s commitment to providing local data processing capabilities for UK organisations with regulatory requirements.

Public sector infrastructure transformation drives data centre demand

The UK Government’s partnership with Google Cloud, announced at Google Cloud Summit London, represents a stark shift and hopes to give the “British taxpayer” a better deal when it comes to technology. 

Currently, public sector organisations face unique challenges in infrastructure modernisation due to legacy contracts and procurement processes. Many government agencies are therefore operating with systems that were implemented decades ago and that have had limited upgrade paths and high switching costs.

“We’re looking to the sector to help shake off the legacy technology that costs the taxpayer an absolute fortune and leaves us vulnerable to outages and to cyberattacks,” says Peter Kyle, Secretary of State for the Department of Science, Innovation and Technology. 

“More than one in four public sector systems run on this ball-and-chain tech, rising to 70% when it comes to police forces right across our country.”

The partnership with Google Cloud includes plans to migrate public sector organisations from legacy systems to cloud infrastructure, with Peter Kyle saying that he hopes this will “liberate public service organisations to use the latest technology and move freely to explore the wider market moving forward”.

Google Cloud’s Extract tool, built with Gemini using Vertex AI, hopes to help councils convert handwritten documents into data within minutes to further support the government in its growth objectives. 

“This is about cutting through bureaucracy, cutting through red tape that's been holding up construction and ultimately to help us build the 1.5 million homes that we've pledged to do over the course of this Parliament,” Peter explains.

“Through agreements like this, we can transition public sector organisations trapped by the ball and chain of legacy services and products to migrate to the cloud. But more than that, that move alone will liberate public service organisations to use the latest technology and move freely to explore the wider market moving forward.”

Charting a sustainable network revolution

Google Cloud’s global network spans multiple regions with redundant connections that enable organisations to deploy AI applications across different geographical locations. This approach requires sophisticated data centre interconnection and content delivery infrastructure.

“We extended our two million mile terrestrial and subsea fibre network by announcing new subsea cables,” Maureen explains. “The UK tech startup ecosystem is absolutely booming and I’m incredibly proud that, since 2023, over 60% of Gen AI startups are Google Cloud customers.”

These are designed to support the bandwidth requirements of distributed AI workloads across data centre facilities. Critically, edge computing capabilities enable organisations to process data closer to end users, which maintain strong security requirements for essential industries, particularly when AI workloads are concerned. 

Additionally, multi-cloud connectivity enables organisations to deploy workloads across different cloud providers whilst maintaining consistent performance and security. This approach drives demand for data centre infrastructure that can support hybrid and multi-cloud architectures.

Energy consumption has also become critical in data centre infrastructure planning as AI workloads require significant computational resources. Google Cloud operates its data centres on 24/7 carbon-free energy as part of its commitment to net zero emissions by 2030 and highlighted in its latest Environmental Report that, in 2024 alone, 2.5GW of new clean energy generation that Google had previously contracted came online.

The sustainability approach extends to cooling systems and facility design that reduce overall energy consumption. At the event, Google Cloud says its data centres incorporate advanced cooling technologies and renewable energy sources to minimise environmental impact whilst maintaining high availability for enterprise workloads.

“After two decades of climate action, we’re driving toward net zero emissions by 2030 and running our data centres on 24/7 carbon-free energy,” Maureen says. “Our goal is to provide the tools and the trust to build boldly.”

The combination of AI workload growth and sustainability requirements is driving innovation in data centre design. Google Cloud suggests that operators must balance computational performance requirements with energy efficiency and environmental compliance in order to succeed.

Peter adds: “All in all, this partnership could see Google invest hundreds of millions of pounds in Britain’s public sector technology, helping to deliver my ambition to bring public services that people use every day and drag them into the 21st century.”