AUS
NZ
UK
CA
USA
IND
Highlights
- NextDC has seen contracted capacity climb sharply as AI workloads flood its data centres.
- Goodman Group has tilted its development pipeline heavily toward power-hungry data infrastructure.
- The scramble for computing capacity is reshaping how the market views Australian infrastructure.
NextDC (ASX:NXT), the country's largest independent operator of data centres, has powered further into focus as demand for the computing capacity behind artificial intelligence keeps surging. Contracted utilisation across its facilities has climbed steeply, driven by a wave of large deals as workloads that require intensive processing pour into the sector. The scramble to build and power the physical backbone of AI has become one of the defining stories on the Australian market, and the companies that own that infrastructure sit right at its centre this winter. What was until recently a quiet niche has moved to the front of the market's attention, and the sheer scale of the capital now flowing into these facilities underlines how seriously the shift is being taken.
The infrastructure behind the AI boom
For all the attention paid to the software and models that make artificial intelligence work, the technology ultimately runs on physical infrastructure. Training and running large AI systems demands enormous amounts of computing power, and that power lives in data centres, vast buildings packed with servers, cooling systems and the electrical capacity to keep them humming. As the appetite for AI has grown, so too has the demand for these facilities, turning a once-obscure corner of infrastructure into a headline theme.
Australia has become a genuine part of that global build-out. Major technology firms are committing large sums to expand their computing footprint in the country, drawn by demand for local capacity and the appeal of keeping data onshore. That has created a powerful tailwind for the businesses that develop, own and operate data centres, and it has reframed how the market thinks about a category of infrastructure once regarded as steady but unexciting.
NextDC and the capacity race
NextDC has positioned itself at the heart of this shift. As the largest independent operator in the market, it develops and runs facilities designed to house the servers of cloud providers, enterprises and, increasingly, AI-focused customers. The recent surge in its contracted capacity reflects how quickly demand has arrived, with large agreements lifting the amount of committed space well beyond where it sat only a short time ago.
The economics of the business rest on filling that capacity with paying customers on long-term contracts, then developing new facilities to meet the next wave of demand. It is a capital-intensive model, requiring heavy upfront spending on land, buildings and power before the revenue flows, but the length and stickiness of the contracts can make it attractive once the facilities are full. The current demand backdrop has put that development engine into overdrive.
Power as the new bottleneck
One of the defining features of the AI era is that access to electricity has become as important as access to land. AI workloads are extraordinarily power-hungry, and securing enough reliable, affordable energy to run large facilities has emerged as a genuine constraint on how fast the sector can grow. Operators that can lock in power supply carry a real advantage, and the race for capacity has quietly become a race for electricity as much as for space.
That dynamic reshapes how the whole sector is understood. A data centre is only as useful as the power feeding it, so the ability to secure grid connections and energy contracts has become a key differentiator. For anyone following the field of ASX AI Stocks, the availability of power has moved from a technical footnote to one of the central questions shaping which operators can capitalise on the surge in demand. ASX AI Stocks
Goodman Group's pivot to data
Goodman Group (ASX:GMG), long known as a developer and manager of industrial property such as warehouses and logistics estates, has tilted its business decisively toward data centres. A large and growing share of its development pipeline is now devoted to these facilities, backed by a substantial bank of secured power across cities around the world. It is a striking reinvention for a business built on sheds and logistics.
The logic behind the pivot is that data centres draw on many of the same strengths as industrial property, namely the ability to secure land, navigate planning and manage complex developments at scale, while offering far stronger demand growth. By redirecting its expertise toward the infrastructure of AI, the group has repositioned itself as a play on computing demand as much as on traditional logistics, a shift the market has watched closely.
From warehouses to server halls
The move from warehouses to server halls is not as large a leap as it first appears. Both require sizeable parcels of well-located land, sophisticated development capability and long-term relationships with major corporate tenants. What data centres add is a far higher value per site and, crucially, that all-important power dimension, which turns a landholding into a genuine platform for AI infrastructure once energy is secured alongside it.
That said, the pivot carries its own demands. Data centres are more complex and capital-hungry than standard industrial buildings, and the pace of technological change adds a layer of risk that warehouses never faced. The market weighs the strength of the demand story against the heavier capital commitment and the need to keep facilities relevant as computing requirements evolve, a balance that will shape how the strategy plays out.
Cooling, chips and the pace of change
The technology inside a data centre is evolving as fast as the demand for it. The specialised processors that power AI generate far more heat than earlier generations of servers, which has pushed operators toward advanced cooling techniques, including liquid systems that were once rare in the industry. Designing facilities that can handle this heat efficiently has become a competitive edge, since wasted energy on cooling eats directly into the economics of running a site.
That pace of change also brings a risk of obsolescence. A facility built for one generation of computing may need reworking to serve the next, and operators must design with enough flexibility to adapt as requirements shift. The businesses that manage this well can keep their facilities relevant and full for longer, while those that fall behind risk owning capacity that no longer suits the workloads customers want to run.
Sovereign capacity and onshore data
A further strand of the story is the growing preference for keeping data within national borders. Governments and large organisations increasingly want sensitive information stored and processed onshore, both for security and to meet regulatory expectations. That appetite for sovereign capacity plays directly to local operators, giving them a source of demand that sits alongside the global technology giants expanding their footprint in the country.
For a domestic operator, that preference can be a meaningful advantage, anchoring demand from customers who specifically want their computing done locally. It adds another layer to the growth story beyond the raw expansion of AI, and it helps explain why the build-out has drawn such attention as a structural theme rather than a fleeting one tied purely to the current wave of enthusiasm around artificial intelligence.
A theme with real weight
What makes this story notable is that it involves some of the larger names on the Australian market rather than speculative minnows. Within the ASX 200, both a major data centre operator and a global property developer now carry meaningful exposure to AI-driven demand, giving the theme a heft that early technology fads often lacked. The infrastructure of artificial intelligence has become a mainstream part of the market's story.
That weight cuts both ways. Because these are substantial businesses, the capital they are committing is significant, and the returns will depend on demand holding up long enough to justify the spending. The market treats the data centre build-out as a durable structural shift rather than a passing craze, but the scale of the investment means the stakes are high if the pace of AI adoption were to slow from its current clip.
What to watch from here
For the data centre operator, the markers worth following are the pace of new contract wins, progress in securing power for future facilities, and how efficiently it can bring new capacity online to meet demand. For the property developer, attention falls on how far its pipeline tilts toward data centres and how successfully it converts its bank of secured power into completed, income-producing facilities.
Underpinning both is the trajectory of AI adoption itself, the single force driving the whole theme. Market participants may assess these names with an eye on whether demand for computing capacity keeps expanding at its recent pace, mindful that the heavy capital being committed today rests on an assumption that the appetite for artificial intelligence has much further to run.
