Nothing Slow And Plenty To Show In Co-Lo; Asanti No Longer Going Solo...

Had a great chat with Rob Garbutt, MD, Asanti Data Centres, following the recent joint venture between my DC/co-lo client Asanti and Lunar Ventures, which is where Rob entered the Asanti building.

Last year, having recorded a few podcasts with the old Asanti team, I was keen to, not only get an update on the new version of the company I knew, but also about the current state of the co-location market and, regardless of all the AI and sovereignty noise, what is actually driving the market – from a customer perspective – in the UK, right now. As well as those driving forces, I was keen to know what the problem areas are, given the demand for capacity but the potential issues in delivering both that capacity and the energy to power it.

What is the current state of the co-lo market in the UK and are there any specific trends emerging?

The UK co-location market is healthy, but it is changing quickly. Demand remains strong, driven by cloud, AI, enterprise resilience and the need for more secure, connected infrastructure. The constraint is no longer just space; it is power, cooling, connectivity and speed of delivery. Recent market commentary points to strong UK demand but limited genuinely deliverable capacity, especially where grid access is constrained.

We are seeing customers think much more carefully about where workloads should sit. Not everything belongs in public cloud, and not everything needs a hyperscale campus. There is growing demand for regional, well-connected, high-resilience co-location that can support hybrid cloud, cyber recovery, edge workloads and higher-density racks.

Has AI really had much of an impact in the UK to date, from your perspective?

AI has had a major impact on sentiment, planning and investment appetite, but the physical impact in the UK is still emerging. A lot of the demand being discussed today is future-facing: training clusters, inference workloads, GPU capacity and high-density deployments.

The immediate effect is that customers, investors and operators are all rethinking power density. Traditional enterprise racks were often designed around relatively modest loads. AI changes that conversation completely. It requires more power per rack, different cooling strategies, stronger network design and much more certainty around energy availability. In the UK, the limiting factor is not ambition; it is whether the right power and infrastructure can be delivered in the right locations.

What do you see as the benefits of the Asanti–Lunar Digital joint venture and what changes can we expect as a result?

The benefit of the JV is that it combines Asanti’s national data centre footprint with Lunar Digital’s hands-on operational, commercial and network experience. The aim is to make Asanti more agile, more efficient and more commercially focused.

The immediate focus is on operational improvement, stronger sales execution, better network capability, improved customer experience and more efficient power usage. We are also looking at how the estate can support the next generation of customer requirements, including higher-density deployments, improved connectivity, cyber recovery use cases and hybrid cloud architectures.

Customers should expect a more responsive business, clearer propositions, better technical engagement and a stronger focus on turning Asanti’s regional footprint into a genuine competitive advantage.

Given the much-documented breaches of major UK and European companies over the past 12 months, especially in retail and automotive, is there now an increasing understanding of the importance of Cyber Recovery as well as Disaster Recovery?

Yes, absolutely. Traditional DR was often built around system failure, site failure or physical disruption. Cyber recovery is different. It assumes the production environment may be compromised, backups may be targeted, credentials may be exposed and recovery needs to happen from a clean, isolated environment.

The high-profile attacks on UK retail and automotive businesses have made boards more aware that cyber incidents are not just IT events; they are operational, financial and reputational events. The JLR attack, for example, was widely reported as causing major production and supply-chain disruption.

That is changing the conversation. Customers are now asking not only “can we recover?” but “can we recover safely, cleanly and quickly if our main environment has been compromised?” Co-location has an important role to play here because it can provide physically separate, secure and resilient infrastructure for immutable backups, clean-room recovery and isolated recovery environments.

Digital sovereignty and data sovereignty are both major talking points in EMEA right now. Are they impacting business decision-making in terms of data storage and choice of applications, cloud vs on-prem/local co-lo etc?

Yes, increasingly. Sovereignty is no longer just a public sector issue. Regulated industries, financial services, healthcare, critical infrastructure and enterprise customers are all asking more detailed questions about where data sits, who has access to it, which jurisdiction applies and how operational control is demonstrated.

That does not mean everyone is moving away from cloud. The future is hybrid. Public cloud remains extremely powerful, but customers are becoming more selective about which workloads go there. Sensitive data, regulated workloads, cyber recovery platforms, latency-sensitive applications and workloads requiring greater control may be better suited to private infrastructure, sovereign cloud, on-prem or local co-location.

For UK co-location providers, that is a real opportunity. We can offer customers UK-based infrastructure, physical control, strong connectivity into public cloud, and a practical route to hybrid architecture without forcing them into a single platform or operating model.

————————————————————————————————————————————-

As impersonator Mike Yarwood used to say when wrapping up his TV show “and this is me”.

Pretty well all of the points Rob highlighted tie in with my own observations of the past 12 months, working with the likes of Goldilock in cybersescurity (the CR angle), Atsign in secure AI (increasing workloads and the importance of securing AI apps at birth, especially in hybrid infrastructures and supply chain scenarios), Element in secure messaging (digital sovereignty in Europe is a huge driver; SoftIron with its OnPrem “cloud” solution sees the same driver) and Aritari in network delivery (those data loads need optimising – securely – more than ever before, as do the server racks in the DC itself; for example, based on monitoring temperatures for performance degradation and moving workloads across servers to minimise that impact).

Rob’s assessment of a hybrid future rings equally true. The reality, as noted in this blog umpteen times, is that the outsourcing nature of cloud services has been around in different guises and with a different name, since IBM and other mainframe companies offered time-sharing in the 70s onwards. So why would it ever stop? The difference is scale; everyone can play the cloud game – modern outsourcing is effectively infinite, IF power availability and capacity/DC expansion can keep up with demand. But then, in IT there’s always an “if”. Would be too easy otherwise 😊.