The power already exists. Buying powered sites is a faster path to capacity than building from the ground up. We upgrade them for modern compute and contract them on long-term leases, turning an underused asset into durable, contracted infrastructure. These sites are smaller and operationally complex, so campus-scale developers look past them — and that's the gap we work in.
The opportunity
Demand is outrunning what the physical world can supply.
Vacancy sits at record lows. New capacity takes years to build, held up by interconnection queues and permitting. The compute is needed now — and the sites that can deliver it are the ones that are already powered.
What we buy
Powered assets with a defined retrofit path.
We target vacant or low-income enterprise, telco and legacy colocation facilities where the existing power exceeds what the building was ever built out to use. Power in place is the whole advantage — it is what lets capacity come online in months instead of years.
What makes a site qualify
- Existing, usable power with documented capacity
- Power headroom beyond the current built-out footprint
- A feasible path to modern cooling and rack density
- Contracted demand that supports the economics
Where we find them
- Fragmented enterprise, telco and legacy colo supply
- Sites below the scale campus buyers chase
- Assets whose complexity narrows the field of buyers
- Entry priced against verified upgrade costs
Who we serve
The middle of the market, where the demand actually is.
The buyer
Mid-market AI teams, enterprises running private models, and platforms in fintech, health and media — the companies that need real capacity but aren't taking a hyperscale campus.
What they need
Energized capacity, live this quarter — not a place in an interconnection queue. Density for modern racks, without permitting or construction risk.
Hyperscalers won't serve a small customer, and a small customer can't finance a campus. We sit in that gap — with power already in place.
How it works
Acquire, stabilize, and recycle the capital.
The same process repeats across separately financed sites. Each one is bought, contracted, upgraded and stabilized — then refinanced, so capital can move to the next.
What sets us apart
Discipline where it's hard to see.
The advantage isn't a single trick — it's how the work is done. Four things compound across every site.
Speed to capacity
Because the power is already in place, capacity comes online in a quarter — not the years a greenfield build takes waiting on interconnection and permitting. We buy the queue-jump, not the queue.
Underwriting the retrofit
Every site is priced against a verified upgrade path. Power headroom and a feasible route to modern cooling and rack density are confirmed before we commit — so the timeline and the economics are real, not hoped for.
Contracted, long-duration cash flow
We secure long-term customer contracts before we build out. That turns an underused asset into durable, contracted infrastructure — value that holds through cycles rather than riding the spot market.
The overlooked middle
These sites are smaller and operationally complex — too small for the campus developers, too intricate for generalist buyers. Fewer bidders means better entry, and a part of the market that receives far less attention than it deserves.
Where we work
Focused on the markets where demand is real.
We concentrate on regions with genuine compute demand and a supply of powered, upgradeable sites — close to the users and systems the capacity serves.
About
A company built by people who have built this industry.
Point is an operating company, governed by operators from data-center design, enterprise technology and AI infrastructure. Its board and leadership have designed data centers, underwritten infrastructure, and run the businesses around them — the unglamorous work that is exactly the point.
Get in touch
Tell us what you're working on.
For capacity, partnership or investment enquiries, send a note. We read everything and reply by email.