AWS just opened a new slice of cloud infrastructure it’s branding as “London,” except most of the actual data centers sit closer to Bristol.
The new availability zone launched this week as part of AWS’s eu-west-2 cloud region and adds capacity for AI and machine learning workloads, including EC2, Trn3, and P6 accelerated instances alongside standard general-purpose cloud compute.
AWS hasn’t disclosed exactly where the physical facilities are located. Availability zones typically sit in clusters spread across a cloud region so a problem at one site doesn’t take down the others nearby, which is likely why AWS is staying vague on specifics even while naming the zone after London.
That naming gap lines up with what’s already public about AWS’s UK cloud buildout. The company is currently developing data centers in Buckinghamshire, north of London, and bought a site in Oxfordshire back in December 2025. It has also been tied to cloud infrastructure projects in Houghton Regis, Hayes, Swindon, and Bracknell, a spread of locations that stretches well beyond London proper.
On the same day, AWS also announced a new Local Zone in Las Vegas, Nevada, extending a different layer of its cloud footprint entirely. Local Zones work as smaller, edge-style extensions of AWS’s cloud, bringing select services closer to specific cities, mainly for applications that need very low latency and can’t afford the delay of routing to a full regional cloud data center. The Las Vegas site runs as a child of the US West Oregon cloud region and supports EC2 instance types C7i, M7i, R7i, and C8gn, along with EBS cloud storage, container services through ECS and EKS, Application Load Balancer, and Direct Connect.
Local Zones aren’t new to AWS’s cloud strategy. The company first introduced them in 2019 and has since rolled out more than 30 worldwide, spanning North America, Europe, Latin America, Africa, and Asia-Pacific. This year alone brought new Local Zones in Istanbul, Hanoi, and Athens, part of a steady push to get cloud infrastructure physically closer to users in cities that don’t have a full AWS region nearby.
Together, the London-area availability zone and the Las Vegas Local Zone reflect two strategies AWS is running in parallel across its global cloud network: building deep regional cloud capacity for AI workloads in some markets, while spreading lighter, latency-focused cloud footholds into cities everywhere else.
