Latos has received planning approval to develop a new GBP £100 million data centre in Stockton-on-Tees, with completion slated for 2027.
Situated on a 1,750-square-metre site at Preston Farm Industrial Estate, the facility is tailored to support artificial intelligence computing at what is termed the ‘neural edge.’ This approach involves establishing smaller, distributed data centres closer to end users, aiming to reduce latency and enhance data-intensive applications such as robotics, autonomous vehicles, and real-time AI systems.
The Stockton site will feature two data halls equipped with Nvidia Blackwell GPUs, a cutting-edge technology optimised for large language models and computer vision tasks. Latos envisions this data centre playing a vital role in boosting local employment and stimulating the digital economy in the Tees Valley region, aligning with its broader five-year plan to develop a total of 40 ‘neural edge’ data centres across the UK by 2030. Future locations include major cities such as Manchester, Birmingham, Leeds, and Glasgow, with the goal of ensuring that no part of Britain is more than 50 miles from an ultra-low latency AI service point.
Unlike traditional, large centralised data centres, Latos’s ‘neural edge’ model aims to bring AI processing closer to users, enabling advanced applications in augmented reality, smart manufacturing, and predictive healthcare. Andy Collin, Managing Director of Latos Data Centres, explained that real-time AI applications like robotics and autonomous transportation require new infrastructure capable of handling demanding workloads efficiently. He highlighted that the neural edge design is more energy-efficient, faster to build, and more cost-effective than conventional data centres.
The Stockton facility will be part of a distributed intelligence network that connects to additional sites planned across the country. It will support AI inference tasks in milliseconds, facilitating more sophisticated real-time AI processes and reducing dependence on centralised cloud processing. A key aspect of the project is its commitment to sovereign data processing, ensuring that data remains within the UK, addressing security and compliance concerns.
Latos pointed out that the Tees Valley region is experiencing rapid growth in its technology and digital sectors, making it an ideal location for this initiative. The area is the fastest-growing in the UK for start-up activity and hosts numerous manufacturing and engineering firms, as well as the government-backed Centre for Process Innovation (CPI). The company expects strong local demand for its services as regional development continues.
As part of its nationwide strategy, Latos plans to build a network of 40 modular, standardised AI processing facilities, designed to be quick to construct and cost-efficient. The company has already announced a 50,400-square-metre hyperscale facility in Cardiff, forming part of its broader rollout aimed at transforming the UK’s AI infrastructure away from traditional cloud models. Collin concluded that this neural edge network represents a fundamental shift that will allow British companies to innovate and thrive using AI technology, rather than relying on outdated centralised systems.
