One thing we are increasingly discussing in the data center industry is transferable skills. Amidst a workforce and skills shortage, it is becoming more apparent that we need to welcome individuals from diverse backgrounds and areas of expertise to drive innovation and expand our capabilities.

That’s why, in this DCD>Talks episode, we hear from Ed Galvin, CEO of DC Byte, as he shares how his experience in commercial property gave him unique insights that have helped elevate data center intelligence to new heights.

Evolution, everywhere, all at once

With a strong background in commercial property charts and surveying, Galvin leveraged his market insights – both differences and similarities – to shape his role in the CBRE data solutions team back in 2015:

“One of the things that really struck me was how we tracked our market – we focused only on Frankfurt, London, Amsterdam, and Paris (FLAP markets). What concerned me was that by analyzing just a small segment and forcing data to fit the model, we risked overlooking key trends.”

One such overlooked trend was the skyrocketing Dublin market, which gradually established itself as a major data center hub – eventually expanding FLAP into FLAP-D nearly nine years later. Galvin reflects:

“What set Dublin apart was that it was largely a self-build market. That’s when the idea for DC Byte took root: we needed to track everything, everywhere, all at once.”

Eight years on, DC Byte has become a leading global data center market intelligence platform, spanning EMEA, APAC, and the Americas. Galvin shares his perspective on the journey:

“One of the greatest challenges is keeping up with the rate of change. Technology defines itself by change, and we’ve certainly embraced that challenge.”

Power and politics

A key challenge facing global data centers is power availability, pushing data centers away from oversaturated Tier I markets and into Tier II and III markets.

Galvin views this shift as a necessity rather than an option, explaining that the industry tends to cluster together, concentrating demands on local utilities. While this results in availability zones where individual buildings can form a single cloud region, modern data centers require at least ten times more power than they once did. Grid operators aren’t equipped to supply this level of power in the necessary timeframes.

“That advantage is now being outweighed by the fact that many of these grid providers are essentially throwing their hands up in frustration. Sometimes, this even involves national governments. To get ahead of national-level moratoriums in certain countries, data center operators need to regionalize their strategy, distributing demand across larger grids to avoid these pinch points.”

Citing examples from the UK and Northern Ireland, Galvin points to the growing number of schemes, like conurbation in the Liverpool and Manchester areas, which have attracted investment from hyperscalers.

According to Galvin, moving away from clustering data centers in major cities and regionalizing demand to prevent grid overload offers additional benefits. With geopolitical and infrastructural uncertainties, such as trade wars and data sovereignty laws, centralizing data storage is risky. The trend toward decentralization could, therefore, be a positive development for the industry by distributing IT infrastructure across countries.

Big brag-a-watt energy

Perhaps it’s time for the Oxford Dictionary to add a new term: “brag-a-watt” data centers, a concept introduced to Galvin by Oliver Jones of AIC and Chayora. This term describes the growing trend in the industry of data center campuses expanding to almost absurd, brag-worthy power levels, reaching into the thousands of megawatts. He adds:

“It’s hard to see the wood for the trees – we know AI is driving significant demand, and it feels like we’re reaching an exponential level of need. On the other hand, new technologies like DeepSeek suggest that perhaps we don’t need as much power.”

Galvin stresses that when these massive, 1,000-megawatt campuses are in the early inquiry phase, it overlooks a critical issue: if all these projects align at once, the data center sector could soon be demanding power equivalent to that of entire countries.

Addressing the impact and potential solutions, Galvin highlights that the industry is bracing for an unprecedented surge in demand driven by AI.

When discussing data center topics that have made mainstream news in 2025, Galvin finds it ironic that markets panic over efficiency gains in the industry, adding that reducing infrastructure costs and power consumption is ultimately beneficial:

“There have been concerns about whether DeepSeek has leveraged the work of previous LLMs, like OpenAI’s models, potentially using more power than necessary. However, what’s more interesting is that DeepSeek is open source, which adds an intriguing new angle.”

That said, security risks – especially around potential Chinese government oversight – are still a concern. The ups and downs in small modular reactor (SMR) stocks reflect broader uncertainty in the market, and it’s still unclear whether large-scale SMR deployment will become a reality. Ultimately, Galvin expects that political responses, particularly in the US, could have a big impact on DeepSeek’s future.

Britain: On the path to AI supremacy?

Another significant development in the UK is the designation of data centers as critical national infrastructure (CNI), a move the government says will provide operators with more support during emergencies.

The announcement of the data center CNI designation aligns with the UK’s ambition to position itself as an AI superpower, making the development of the necessary infrastructure a top priority. The main thing Galvin hopes for here is faster planning:

“One of the challenges is that there’s already an excess of demand for data center space, so most of the committed space under construction, and all the live capacity within the market, is currently pre-sold. We need to build more capacity quickly if we are going to host all those AI training servers and keep ourselves ahead.”

Galvin leaves us with this thought: if we want to capitalize on the AI wave, we need to be prepared to ride it when it comes.

To get expert insights from Galvin on global data center construction trends and scaling digital infrastructure in the AI era, watch the latest DCD>Talks episode here.