- The rise of AI inference computing is driving demand for smaller, edge data centers located near major population centers.
- Adaptive reuse of existing industrial and mission-critical facilities is reemerging as a viable — and sometimes faster — alternative to ground-up construction.
- While skepticism persists, successful examples like Bit Digital’s North Carolina factory conversion suggest this trend may gain traction as AI infrastructure demands intensify.
AI Is Changing The Rules
After years of favoring new builds over retrofits, data center developers are rethinking their approach, reports Bisnow. The shift is driven by the growing need for AI inference computing — the stage where users interact with AI models — which requires smaller, distributed facilities close to end users.
Back In Play
Adaptive reuse once played a key role in early data center development, with industrial and telecom buildings repurposed across the country. But over the past decade, ground-up construction became the norm, as hyperscale data centers grew in size and complexity.
Get Smarter about what matters in CRE
Stay ahead of trends in commercial real estate with CRE Daily – the free newsletter delivering everything you need to start your day in just 5-minutes
Why Now
AI inference, unlike training models which demand centralized computing power, benefits from decentralization. That’s driving demand for edge data centers in urban areas, often where old warehouses or industrial sites can be reused.
According to TYLin SVP Kirk Mettam, “Adaptive reuse is not that popular of a term, but I think it’s going to become more and more pertinent as the edge market evolves.”
The Bit Digital Blueprint
AI computing firm Bit Digital is betting on this approach. It’s converting a former factory in Madison, North Carolina, into a 24-megawatt data center — the first phase of a larger 96 MW campus. Its proximity to Charlotte and Northern Virginia makes it ideal for edge AI workloads.
CEO Sam Tabar said existing power and fiber made retrofitting faster and cheaper than starting from scratch.
Skepticism Remains
Despite renewed interest, many developers still view retrofits as costly and inefficient, especially when dealing with buildings not designed to support the weight, cooling, and ceiling clearance requirements of modern data centers.
“Once a company does an industrial-to-data center conversion, that will be the last one they do,” said JLL’s Carl Beardsley earlier this year.
What’s Next
As AI demand evolves, so will the data center form factor. Industry leaders urge developers to rethink assumptions and assess when adaptive reuse works best, especially for AI-driven edge deployments.
“There are many underutilized assets on the market right now,” Mettam said. “As we watch how the needs of the data center industry are evolving, we can find situations where it rightsizes.”
Bottom Line
Adaptive reuse won’t replace ground-up builds — but with AI’s rising demand for edge infrastructure, developers may need to dust off this old playbook.
