Data centers have long been cemented as the backbone of the global digital economy, experiencing impressive levels of rapid growth thanks to accelerated demand for advanced compute networks.

The rise of AI has been central in intensifying this expansion, driving significant investment in the design and deployment of future-ready facilities equipped to handle increasingly complex computational needs, rising densities, and advanced cooling and connectivity requirements.

As key stakeholders across the data center ecosystem race to meet these rising demands, global data center expansion is advancing into emerging markets across the world, bringing an equal number of promising opportunities and complex challenges to navigate.

Against this backdrop, in a recent DCD>Talks episode Steven Lim, senior vice president of marketing and GTM strategy at NTT Global Data Centers, shares his view of global expansion across the data center industry, the key drivers behind this growth, and where we can expect these trends to continue gaining traction.

Compounding growth

“AI, as we all know, is the hot topic, and it’s driving awareness of the data center space like never before,” says Lim. “The growth we’re seeing is significant. Huge deployments are happening, mainly in the US, but quickly reaching into other markets as well.”

While AI may be dominating the headlines, Lim is quick to emphasize that this doesn’t mean traditional cloud or hyperscale workloads are slowing down:

“I think traditional hyperscale business is starting to get overlooked a little bit. Cloud services and enterprises are still growing at pace. By some early estimates, the data center industry was growing at around 20 to 24 percent CAGR, and AI was initially an additional ten to 15 percent on top of that. Over the last six to eight months, that number may have crept up even higher.”

This reality reveals that AI and traditional hyperscale are in fact evolving in parallel as two powerful drivers, creating a compounding growth that’s driving explosive development across the data center industry at scale on a global stage.

The return of the enterprise

In the face of impressive AI growth and increasing recognition of its value, enterprise data centers have similarly been starved of the limelight. Yet, this critical section of the market isn’t going anywhere. Lim explains:

“Over the last few years, enterprise clients in some ways have been shut out of the business. Clients needing three, six, or even ten to 12 megawatts often saw those deals go to hyperscalers or early AI developers who wanted to grab everything available. As a colocation provider, that’s hard to pass up.”

This dynamic, Lim explains, created pent-up demand from enterprises, which is now beginning to surface as these key players make aggressive moves to get back into the market and meet their specific demands.

As these organizations re-enter, they’re following the same trends seen in hyperscale and AI deployments, particularly around density. “From enterprise, to hyperscale, to AI, everything is becoming much denser,” says Lim. “Managing this density is very different from what we did ten years ago.”

A key shift driven by rising densities is the transition from predominantly air-cooled facilities to the necessity for liquid cooling inside data centers. This trend, once exclusive to hyperscale and AI applications, is now filtering down to the enterprise level at pace.

“We’re seeing enterprises come to us saying, ‘I need six or 12 megawatts – but it needs to be water-cooled as well.’ Ultimately, the trend now is for enterprises to apply the lessons learned from hyperscale and AI to their own environments.”

Established and emerging markets

While it’s clear that the industry is experiencing growth across multiple fronts, it’s no secret that AI remains the dominant driver in the global arena.

“One report revealed that a little over $30 billion was spent on AI deployments in 2024 alone,” says Lim. “That’s a massive number. If we go back five or six years, a billion-dollar investment would have been considered huge.”

Adding to this picture, a report from McKinsey & Company projected that by 2030, 70 percent of data center growth will be driven by AI.

“The US is still the dominant player in this AI development that we’re seeing,” adds Lim. “The challenge is how we combine all these layers of growth – incorporating enterprise and hyperscale demands too – over the next three to five years, that’s what we’ll see play out primarily in the US.”

But this isn’t just a US story. Expansion across emerging markets, especially India, is quickly gaining ground and grabbing attention.

“There’s a lot of development happening in India because the demand is so high,” says Lim. “One of the unique things about the country is that it’s a digital-first market. Traditional regions like the US or EMEA (Europe, the Middle East, and Africa) evolved from legacy wired infrastructure, but India didn’t have that, so it’s leapfrogged straight into digital-first connectivity.”

As a relatively lower-cost market where land and construction opportunities are more readily available than many other areas, plus a vast population and growing digital ecosystem, India looks to be an attractive investment, well-positioned to grow into a data center stronghold.

But it’s not all plain sailing, and Lim highlights the core challenges that remain in this, and multiple other emerging markets alike: “One of the key issues will be power stability and regulation. It’s still a bit of a wild west compared to the US or EMEA, so how that’s managed will be crucial as growth accelerates.”

APAC, EMEA, and FLAP

The APAC (Asia-Pacific) region is experiencing its own diversification. Traditionally, growth was centered in Hong Kong and Singapore, but these markets have limited land availability, driving expansion into areas such as Johor, Bangkok, Indonesia, and Malaysia.

“What will drive continued growth is land and power availability,” reiterates Lim. “Plus, whether the right networking and connectivity infrastructure exists to move data efficiently in and out. I suspect we’ll see these emerging markets mature into Tier 1 or Tier 2 hubs to meet the region’s massive demand.”

Shifting focus to EMEA, Lim highlights further opportunity and complexity: “EMEA is interesting because it’s such a diverse region with so many countries,” he says. “We’ll continue to see growth in the core FLAP markets – Frankfurt, London, Amsterdam, and Paris – but also more distributed expansion.

It’s clear that AI expansion, initially gaining momentum in the US, is quickly spreading into key European markets, with many regions already poised to capture the opportunities. Lim expects the demand wave in this region to occur approximately 18 to 24 months behind that in the US.

While this key trend is well underway, two key inhibitors still demand close attention: sustainability and data sovereignty.

“Building and delivering sustainably is essential – Europe is leading the charge on that,” says Lim. “The other issue is data sovereignty. It’s much more prevalent across Europe. Everyone’s familiar with GDPR, and AI will have to be addressed within that framework. How we manage, store, and govern data will be critical.”

Overcoming the core constraints

Despite immense opportunity, the industry also faces three primary constraints to fast, continuous growth across diverse regions: power, land, and capital.

“I’d say number one is power,” states Lim. “Is there enough to support this level of growth? AI deployments are at a scale we’ve never seen before. We used to talk about rack densities of three or four kilowatts – now we’re talking 100 kilowatts or more. The latest Nvidia chips could push that to 500 kilowatts, even a megawatt per rack, which is just mind-boggling.”

The second constraint is land. “Land and power are intimately tied,” he continues. “You can’t look at one without the other. You need the right location, with the right connectivity and power availability for a site to be viable.”

And finally, there’s the issue of capital. The cost to build keeps rising – from construction materials to equipment. The key question therefore becomes: is there enough funding to capture this business and build at the pace required to support the rate of technological advancement?

A look to the future

As AI reshapes the digital landscape, the data center industry is entering its most transformative phase yet. The convergence of hyperscale, enterprise, and AI-driven demand is pushing infrastructure design, investment, and sustainability to new limits, and unlocking unprecedented global opportunity.

“AI is changing everything – how we build, where we build, and how we operate,” concludes Lim. “But the fundamentals remain the same: power, land, and capital are the foundations of growth. The winners will be those who can balance all three, while building sustainably and globally.”

With AI acting as both catalyst and challenge, key industry players must turn focus to ensuring that innovation, collaboration, and sustainability stay at the heart of this global transformation.

To hear more about Steven Lim’s view of the most promising global data center trends, watch the full DCD>Talks episode here.