Artificial intelligence is no longer reshaping software alone. It is transforming the physical infrastructure that powers the digital economy and creating new opportunities for investors, developers and landowners. As a new generation of cloud providers emerges, the rules governing data centre investment are beginning to change.

For more than a decade, the data centre market has been driven by a relatively small group of technology giants. Amazon Web Services, Microsoft Azure and Google Cloud – the so-called hyperscalers – have defined the industry's scale, seeking vast sites with access to hundreds of megawatts of power to support large-scale digital infrastructure.

That model remains essential. However, the rapid expansion of artificial intelligence is creating a new layer of demand, one that could significantly reshape how digital infrastructure assets are valued.

Introducing neoclouds

Unlike traditional hyperscalers, neocloud providers focus almost exclusively on delivering high-performance computing for AI workloads through GPU-as-a-Service models. Instead of investing in expensive hardware and managing their own infrastructure, companies can access the GPU capacity they need through specialised providers, allowing them to scale AI projects more quickly and efficiently.

The timing is no coincidence. Demand for AI computing capacity is growing far faster than new infrastructure can be delivered. Businesses across industries, research institutions and technology companies are accelerating AI adoption, yet many cannot afford to wait years for new hyperscale campuses to become operational.

Industry forecasts underline the scale of this transformation. Gartner estimates that neocloud providers could account for around 20% of the global AI cloud infrastructure market by 2030, representing a market worth approximately $267 billion. The consultancy also predicts that by 2029, AI workloads will consume around half of the world's cloud computing capacity, compared with less than 10% today.

These figures point to more than technological evolution. They signal a structural shift in digital infrastructure. Perhaps the most significant consequence, however, lies beyond technology itself.

For years, data centre development has largely focused on hyperscale projects requiring well over 100 megawatts of available power. These projects will continue to play a critical role. Yet across Europe, securing large energy allocations, navigating lengthy permitting processes and identifying suitable land have become increasingly challenging.

How neoclouds are changing the game

Specialised AI operators often require considerably less capacity than hyperscale campuses. As a result, sites that may previously have been overlooked are becoming commercially viable. Land with access to 30, 50 or 80 megawatts of power – provided it also offers strong connectivity, grid access, planning certainty and proximity to major digital networks – may now represent attractive opportunities. In other words, artificial intelligence is changing the way digital infrastructure assets are valued.

Location remains important, but it is increasingly being matched by other critical factors: immediate access to power, network connectivity, expansion potential, proximity to peering hubs and, perhaps most importantly, speed of execution.

This trend also creates opportunities for countries such as Portugal. Its Atlantic location, growing network of international submarine cables and expanding digital connectivity position the country favourably within Europe's digital infrastructure landscape. While Portugal may not always offer the vast energy capacity required by every hyperscale development, it does possess many sites capable of supporting the next generation of specialised AI infrastructure.

That does not mean every plot of land with available power is suitable for a data centre. These remain highly complex developments requiring technical expertise, regulatory certainty, grid availability and experienced partners.

What does this demand mean for investors?

For investors, developers and landowners, this requires a different way of assessing opportunities. The future of digital infrastructure will not be shaped solely by hyperscalers. A new generation of AI-focused operators is emerging, bringing different requirements, faster decision-making and a more flexible approach to site selection.

As artificial intelligence continues to accelerate, the competitiveness of an asset may depend less on its size and more on the combination of available power, connectivity, planning readiness and the ability to respond quickly to market demand.

Neoclouds are only one manifestation of this transformation. But they also offer an early indication of how the next generation of data centre investment is likely to evolve, and of the new opportunities waiting to be recognised by those prepared to look beyond the traditional hyperscale model.

Whether you're expanding your portfolio or entering the commercial real estate market, Athena Commercial provides expert guidance in identifying the most suitable opportunities.

info@athenaadvisers.com

+351 917 702 022