Italy is drawing another nine-figure bet on AI infrastructure.
Greenfield and Finsbury Infrastructure plan to develop a 36MW data center east of Milan, with up to €360 million expected to be invested in the project and the surrounding region over the next three years. Planned capacity is intended for AI, cloud, and enterprise workloads.
A key piece of the build is already in place, and the partnership also opens a new market for one of the developers.
Power deal draws on AI infrastructure expertise
Greenfield has secured 80MW of power capacity for the site, more than twice the planned IT load.
It brings an established energy-development business to the project. It reports 66 active projects and 5GW in development, with in-house teams covering land, grid, design, legal, and planning. Its data center work also incorporates energy-efficient design, cooling, and renewable power options.
Finsbury supplies capital and expertise geared toward AI-scale infrastructure. Its data center model uses high-density designs with advanced cooling and targets sites near high-voltage power and dark-fiber networks. Its larger campuses are designed for more than 100MW of compute capacity, although those specifications have not been announced for this partnership.
Finsbury enters Italy with more projects in sight
Milan is Finsbury’s first project in the country. Finance Director Peter Connolly said the company expects additional investments as the market takes on a larger role in Europe’s digital economy.
The data center infrastructure developer already has activity in Dublin, London, and Madrid, extending its footprint into another major European market.
Greenfield adds local experience to the partnership. Its data center portfolio includes developments across Lombardy, Piedmont, Lazio, and Puglia.
Channel opportunity depends on what fills the facility
MSPs, VARs, and systems integrators serving customers in Italy and nearby European markets could eventually gain another regional option for AI workloads if the project comes online as planned.
New deployments around the facility could generate services work beyond the underlying compute. Moving customer workloads into Milan may call for hybrid-cloud integration, migration support, managed security, and governance around what data and systems move there.
Milan could also become another location for advisers to weigh when customers compare local infrastructure providers. Capacity, proximity to customer operations, contract flexibility, and dependence on a single provider could all influence whether the site fits a particular deployment.
Before steering customer AI deployments toward the site, partners should verify:
- How much compute will be available when the facility opens
- What hardware will support customer workloads
- When production services are expected to become available
- Whether capacity can be purchased through partners or only through direct agreements
Partner access will shape the commercial upside. A channel-friendly sales model could open direct services revenue around the facility. If capacity is sold mainly to large infrastructure buyers, MSPs and integrators may see more opportunity supporting customers that use it than selling the capacity itself.
Also read: Gimlet Labs is scaling toward hundreds of megawatts of managed AI infrastructure after raising $300 million at a $3 billion valuation.