Real estate and Artificial Intelligence clusters
We explore the UK's thriving innovation hubs and how real estate should respond to Artificial Intelligence advancements.
5 minutes to read
Artificial intelligence (AI), particularly generative AI, is the hot topic of 2023.
Interest in this upending technology goes beyond curiosity and is now turning towards commerciality. There has been £1.6bn invested in UK AI companies alone during the first six months of 2023, across almost 300 deals. This has led to AI surpassing fintech as the leading UK tech vertical for equity investment.
The rapid adoption of models such as GPT-4, a three-fold increase in UK job postings requesting Artificial Intelligence skills over the past decade and a ten-fold increase in UK LinkedIn members with AI skills since the beginning of 2016 all illustrate AI’s astonishing growth trajectory.
UK Artificial Intelligence hub
This is good news for the UK which is an established world leader in the AI space. Indeed, there are more than 120,000 people employed in 3,000 AI companies within the UK. The UK also ranks third globally for both AI publication citations per capita and equity investment, and is home to leading companies such as DeepMind, Graphcore, Stability AI and Darktrace alongside leading research institutions such as the Alan Turing Institute.
The sector's growth will be further accelerated by government action, an evolving regulatory landscape, robust investment, rapid adoption across various sectors, and further technological advancements.
Government support
Government support for the AI sector is increasing as witnessed by the recently launched National AI Strategy which provides a blueprint for the safe and responsible development of AI, the formation of an expert task force, and a commitment to spend £1bn over five years on AI and supercomputing.
In June, Rishi Sunak said Britain would host the “first global summit on Artificial Intelligence” this autumn.
Such activity will continue to translate into real estate demand, with high-profile companies like OpenAI choosing the UK as the location for their first international office.
Over the past decade, there has been an average of 200 new AI companies founded each year in the UK. Assuming growth continues at this pace, there will be an additional 2,000 companies looking for space over the next ten years.
Against this positive, expansionary backdrop, this article outlines:
- Key emerging and established AI clusters within the UK
- A checklist of the key real estate considerations for those seeking to attract AI occupiers
- Innovation clusters
In analysing leading indicators such as business counts, number of graduates, equity investment, availability of talent and university rankings specific to AI, we have constructed an index to determine the UK’s leading AI clusters.
London is the leading UK cluster for AI. It is home to just over 2,000 AI companies and leading research institutions, such as the Alan Turing Institute. It is of note that the boroughs of Camden, Westminster and City of London are home to 40% of London’s AI companies.
Next in line are Edinburgh, Cambridge, and Manchester. The University of Edinburgh is one of Europe’s largest centres for AI research, with the greatest UK higher-education student enrolment in AI-related courses. The city is also home to 87 AI companies and 2,500 AI professionals.
Manchester ranked highly for the number of professionals with AI skills, being second only to London, while Cambridge-based AI companies have received significant venture capital funding.
Companies that have received funding in recent times include Healx and Darktrace. Cambridge University is also ranked second in the UK and sixth in the world for computer science, whilst Cambridge sits squarely at the intersection of AI and other sectors, such as life sciences.
Top 10 UK AI clusters
Real estate considerations
Companies focused on AI have some unique real estate needs. Firstly, flexibility is key, given that 80% of companies are either SMEs and/or at seed, start-up, or scale-up stage. Proximity to talent and like-minded companies in a collaborative ecosystem is also high on the checklist as well as buildings and neighbourhoods that provide an exemplar employee-experience.
Building and spec design
Infrastructure such as a reliable power supply, high-speed connectivity, adequate cooling and data storage are clearly vital. Power and data contingency options are needed in case the primary sources fail. AI companies will want to know about data storage in the area or building and the ability to scale-up capacity.
As generative AI continues to rapidly advance, data centres will have to adapt. AI applications consume significant amounts of power, utilising high-performance processors, and posing implications for energy use in data centres.
Power-usage and workloads
AI applications will typically use around 30kW of power per rack, a significant jump from the more traditional data centre rack, which utilises 8-10kW. These servers also require higher power density, leading to greater cooling demands and undoubtedly furthering power demands.
Generative AI workloads can be classified into two categories: Training and Inference, with both outlaying a different set of requirements. Training workloads require more computing power, via advanced GPUs, whilst Inference carries a lower computing load and instead relies on a low-latency connection. We could see specialised facilities built focussed on both sides of generative AI:
- High-powered, remote facilities to handle Training
- Low-powered, localised facilities to handle Inference
Building design must consider the needs of people using the space and the tasks they are engaged in. This could include designing for neurodiversity, given the fact that large tech companies associated with AI recognise the importance of neurodiverse individuals within their workforce. It could also mean the provision of collaborative spaces as well as quiet zones for concentrated work. Certain companies may require lab space for testing AI robotics and hardware prototypes.
More broadly, AI is supporting transformation in different sectors, with implications for the real estate they occupy. In life sciences, for example the R&D process is being transformed by AI, resulting in changes to lab design and the relative proportion of office vs lab space that life science occupiers require.
Real estate opportunity
The AI sector is set for rapid growth, with the UK home to a number of emerging and established clusters. This creates an opportunity for those on the supply-side who can map and monitor the occupier landscape and met their real estate needs.
Subscribe for more
Get exclusive market analysis, news and data from our research team, straight to your inbox.
Subscribe here