Asia-Pacific emerging as the fastest-growing region for data centers
Asia-Pacific’s digital infrastructure is scaling fast, and for good reason. If you’re thinking about where the most aggressive data growth will happen next, this is where your attention should be. Countries like India, Malaysia, Vietnam, and Australia are shaping up into serious data center competitors, not just fringe markets. Singapore is still a major player, even with its land constraints.
Domestic demand across APAC is rising fast. Government policy is aligned with that demand, and land availability is being unlocked in key corridors. In India, hyperscale providers are responding to widespread cloud adoption. Malaysia is winning spillover demand from Singapore, offering lower costs and strong network capabilities. Vietnam is in early stages, but momentum is real. Meanwhile, Australia continues to attract capacity investments thanks to its stable grid and connectivity to major Asian economies.
For companies with global infrastructure footprints and speed-to-market goals, these are not just options, they’re priorities. Infrastructure developers aren’t thinking in terms of single country footprints anymore. They’re deploying regional portfolios. That means one operator may serve traffic from Vietnam, support storage from Malaysia, and control compute from Singapore or Sydney. This cross-border strategy is becoming the norm.
Also consider this: Southeast Asian governments aren’t waiting for the market to lead. They’re setting targets for cloud services and digital trade. That changes the risk profile for investors, reducing policy uncertainty and increasing clarity on long-term commitments. It’s structured opportunity. When regulation aligns with business objectives and digital demand is already scaling, you get acceleration.
This is the most dynamic growth region in digital infrastructure right now. APAC isn’t catching up, it’s defining the pace.
Alex Saez, Partner of Data Centers at Cundall, put it plainly: policy choices, local market maturity, and smart land and energy planning are shifting the investment map. It’s no longer just about chasing demand, it’s about betting on the regions that are building their ecosystems to support that demand at scale.
UK faces an escalating skills shortage in its burgeoning AI and data center projects
The UK wants to lead the AI race, and the ambition is real. Investment zones are forming, funds are flowing, and the government is backing high-stakes projects to build data infrastructure across the country. But here’s the problem: that infrastructure doesn’t build itself, and right now, the local labor market isn’t keeping up.
There’s a widening gap between demand and the supply of qualified engineers and construction professionals. Data centers being planned today are far more complex than those of five years ago. They support dense AI compute, dynamic workloads, custom power architecture, and tight sustainability standards. You can’t scale that kind of infrastructure without the right talent, and today, the UK doesn’t have enough of it.
Skills are a bottleneck. Andrew Livesey, Partner for Data Centers at Cundall, made this pretty clear: if the UK is serious about becoming an AI superpower, it needs to fix its workforce pipeline. A narrow talent pool slows down critical builds. He’s advocating for wider participation, getting more diverse, younger, and non-traditional candidates into engineering routes, fast.
That means more apprenticeships, stronger training pathways, and better alignment between industry needs and education. This isn’t just about junior talent, it’s about building a workforce that reflects the scale and complexity of what’s ahead. Livesey points to initiatives like The Land Collective, which is actively helping people who wouldn’t normally enter engineering see it as a viable, valuable path.
The good news: momentum is shifting. Major data center operators and engineering contractors are already boosting their apprenticeship programs. Industry bodies are pushing for deeper inclusion and broader hiring strategies. This matters, not just for talent diversity, but for delivery speed and operational continuity.
If you lead an infrastructure-heavy business in the UK, this is not a long-term problem, it’s today’s risk factor. Training takes time. So if you want to build or expand data capabilities locally, your organisation should already be investing in the talent side, not just the hardware. And if you’re not influencing education and apprenticeship coalitions today, you’ll be playing catch-up tomorrow.
AI-driven computing demands are revolutionizing data center power infrastructure design
AI isn’t just pushing more data through data centers, it’s changing how those centers function at their core. The next generation of workloads demands new infrastructure. The conversation used to circle around cooling challenges. That’s changing quickly. The focus now is on power. More specifically, how to supply it fast, efficiently, and with stability, at a scale that traditional systems weren’t built to handle.
GPU clusters, which drive most AI compute today, don’t behave like conventional systems. Their power usage isn’t steady, it can spike without warning. One moment they’re sitting at 20% load, and a few processing cycles later they’re pushing 160%. That level of volatility stresses everything from rack-level distribution to campus-wide grid connections. And the numbers are only going up. Rack densities now exceed 300kW, and there’s already a competitive roadmap aiming at 2MW per rack.
Jamie Cameron, Partner for Data Centers at Cundall, put it bluntly: “How do we power this?” That’s now the defining question. AI builds require rethinking power architecture from the rack up to the grid. Liquid cooling is largely solved. Electrical systems are next.
The solutions are emerging. 800-volt direct current (DC) distribution systems are gaining traction, they move power more effectively at high density. Supercapacitors are now being designed into racks to manage short, sharp load peaks. Battery Energy Storage Systems (BESS) are being introduced at the campus level to buffer multi-megawatt step loads without hammering the wider grid. These aren’t theoretical ideas. Proof-of-concept deployments are already underway with developers and utilities collaborating closely.
Standardized designs based on these new configurations will become the framework for future hyperscale builds. You won’t see them in wide adoption by 2026, but by then, leaders in the space will be deep into pilot phases. They’ll have real data, field-tested performance, and a head start on grid negotiations and energy stability planning.
For executive teams making infrastructure bets, the takeaway’s simple: focus your power strategy around both infrastructure and technology innovation. If your facility design is still anchored to traditional loads or cooling-first approaches, you’re already late. The AI layer changes the equation. You’ll need both resilience and flexibility engineered into every part of the electrical chain, from rack to grid.
Regional disparities in resource availability and policy frameworks shape the landscape of data center growth
The global data center market is fragmenting, fast. Access to power, land, and aligned regulation now determines which countries attract new builds and which fall behind. Growth is not distributed evenly. Some regions are scaling up, securing projects through clear digital infrastructure strategies. Others face saturation, grid instability, or regulatory obstacles that stall development.
As complexity increases, developers and hyperscale operators are shifting from single-market strategies to regional portfolios. They’re no longer watching one country, they’re evaluating clusters, stacking cities, utility models, and policy trends across multiple geographies to build resilient, flexible infrastructure footprints. That change reflects a deeper reality: not every market is fit to support high-density, AI-ready assets.
In Europe, there’s divergence. A few markets are holding or gaining ground, those with smart land-use policy and high grid reliability. Others, constrained by rigid zoning, insufficient power, or political bottlenecks, are seeing project slowdowns or redirection. That regional imbalance drives smarter capital deployment. It forces operators to align investment not only with demand but with execution conditions on the ground.
Barbara Sacha, Partner and Data Centre Sector Lead at Cundall, pointed out that location decisions are increasingly shaped by the mix of AI compute growth and practical infrastructure variables, land, power, and permit velocity. Strong demand alone no longer guarantees a data center build. The operating environment must support advanced deployment schedules, high rack density, and long-term energy reliability.
For executive leadership, this shift calls for more operational awareness and strategic independence. Relying solely on legacy tier-one markets may miss faster-moving corridors in less saturated areas. Global operators need united infrastructure and policy teams to track and respond to these shifts. And market-entry decisions must be based not just on GDP or user density, but on true build-readiness, power, land access, and forward-leaning governance.
The future of data centers isn’t about geography, it’s about adaptability to changing conditions. The markets that offer deployment alignment with AI and cloud scale will own the next phase of high-performance infrastructure expansion.
Key takeaways for leaders
- APAC accelerates data center dominance: Leaders should prioritize investment in Asia-Pacific markets like India, Malaysia, and Vietnam, where digital demand, land availability, and supportive government policies are driving hyperscale growth at unmatched speed.
- UK faces structural talent gap: Decision-makers must invest in workforce development, via apprenticeships and inclusive pipelines, to avoid delays in scaling infrastructure tied to AI and cloud initiatives across the UK.
- AI disrupts power planning strategy: Leaders should reassess data center power architectures to handle high-density, volatile AI compute loads, incorporating emerging standards like 800V DC distribution and battery energy storage at scale.
- Location strategy needs a policy lens: Executives managing global infrastructure should evaluate markets not just for demand, but for execution readiness, specifically land access, power availability, and regulatory efficiency, to mitigate deployment risk.
