HSCALE, the hyperscale data centre firm backed by Bain Capital, has announced the closing of its second large-scale data centre campus in Northwest Milan. The transaction brings HSCALE’s total committed power capacity in the Milan metropolitan area to 250MW, with ready-for-service dates in 2028.
The combined Milan investment represents over €2 billion of capital deployment in the region, with pre-construction and procurement activities already underway to ensure delivery in line with customer demand.
Doubling Down on Milan
Both HSCALE campuses are located within Northwest Milan (Settimo): one of the most constrained and strategically important geographies in the market. All major hyperscalers are already present and growing in this area, making it a primary deployment zone for cloud infrastructure in Southern Europe.
Both sites are fully owned with power committed with key development milestones already achieved.
This level of site control and delivery certainty sets HSCALE apart in a market where securing land, power and permits can take a considerable amount of time. With both sites advancing through pre-construction, HSCALE is positioned to deliver capacity at scale and on schedule.
Built for the Future
HSCALE’s Milan campuses feature a design philosophy refined over 12 months of engineering work to align with hyperscaler specifications. The base design allows customers to move between different cooling methodologies: air-cooled, direct liquid-cooled or hybrid configurations with minimal impact on customer spend.
This flexibility matters, as high-density AI workloads increasingly require high-performance liquid cooling, while traditional networking, compute and storage systems continue to grow and still require a mixed combination of cooling methods. HSCALE’s approach means customers can deploy today and adapt as their infrastructure needs evolve, without redesigning their facilities.
A vision for the region
“We designed HSCALE's Milan campuses around a simple principle, the building should never be the bottleneck. Our base design is liquid-cooled first, built for the most demanding hyperscale and AI workloads, and can pivot to air-cooled traditional deployments in the same physical structure. No redesign, no additional capex. We design and build like this because we understand the long-term commitments our customers must make.”