Summary
- SWI Group reported €4.4bn of assets and €2.3bn adjusted NAV for the first half.
- Its European AiOnX platform has about 2.3GW of planned data centre capacity.
- The strategy combines powered land, data centres, and AI compute rather than treating the assets as standalone property.
SWI Group has put its European data centre development pipeline at approximately 2.3GW as the investment company accelerates a strategy combining powered land, data centres, and AI compute infrastructure.
SWI Group said in its first-half results that total assets reached €4.4bn at the end of June, while adjusted net asset value rose 53% from the end of 2025 to €2.3bn.
The group’s European digital infrastructure activities are centred on AiOnX, which it says has about 2.3GW of planned capacity, including a campus leased to a hyperscale customer. Its wider infrastructure portfolio also includes SWI Digital in the United States, giving the group exposure to roughly 4GW of power capacity across the two regions.
The numbers illustrate a broader shift in data centre investment towards assets where the scarcity lies in secured power and development rights rather than the building alone. Land that can obtain a large grid connection has become a strategic input for AI infrastructure, particularly as proposed campuses move into hundreds of megawatts.
Power is becoming the underlying asset
AiOnX’s pipeline spans several European markets and has been assembled around sites intended for hyperscale development. Earlier material from the platform identified projects in Ireland, Denmark, Spain, Italy, and the UK.
One of its Dublin phases is designed around a 16MW first stage with hyperscale rental income expected as capacity becomes operational. SWI has also expanded its strategy into AI compute through investment in GPU infrastructure provider Polarise, tying the data centre development platform more directly to the workloads expected to occupy it.
That vertical approach reflects the changing economics of the sector. A developer that controls land and a grid connection but has no customer still faces financing and demand risk. A compute provider without power and suitable facilities faces the opposite constraint. Combining the layers can create more control over delivery, but also increases capital requirements and execution exposure.
SWI’s results say the company is moving further in that direction. The group has described digital infrastructure as a core investment area and has been reducing emphasis on non-core assets while increasing exposure to powered sites and compute.
Pipeline size needs delivery discipline
A 2.3GW planned portfolio should not be read as operational capacity. Development pipelines contain projects at different stages of planning, grid connection, financing, leasing, and construction.
The distinction is particularly important in the current market because power applications can be far larger than the amount of capacity likely to be built on the original timetable. Large campuses are commonly phased, allowing operators and investors to align capital expenditure with contracted demand and the delivery of network infrastructure.
AiOnX’s value will therefore depend less on the headline gigawatt total than on how much of that power is secured, when connections can be energised, and whether customer commitments support financing.
The presence of a hyperscale tenant on one campus gives SWI a route to contracted demand, while its investment in AI compute provides another potential source of utilisation. Both strategies still depend on construction delivery, hardware availability, cooling design, and electricity infrastructure reaching the sites on schedule.
The group’s results show how digital infrastructure investors are increasingly moving beyond conventional real estate. In an AI market where usable megawatts are scarce, the commercial stack increasingly begins with the grid connection and extends through the facility to the compute installed inside it.
