Complete campus development for single anchor tenants — secured power, engineering for dense AI loads, and schedules measured in months, not years.
One campus, one anchor tenant, one accountable developer. We acquire and entitle the land, secure the power, design and deliver the buildings, and operate the campus — our tenants deploy compute. Right-sized means matching each campus to what the site, the grid, and the community can support. Sometimes that's 20 MW; where we generate the power ourselves, it's 400 MW.
Interconnection, withdrawal capacity, and generation planning are validated before a campus enters our active portfolio. Power is engineering, not a press release.
Proprietary technology runs our land qualification, diligence, underwriting, and predictive siting. Once the tenant signs, we pay for the substations, feeders, and protection upgrades — never the utility or its members.
Every campus advances through a staged development gate process — site control, power, design, financing, construction — with institutional-grade diligence at each gate.
460 MW of critical IT across Georgia, Texas, and Ohio — with 240 MW in commercial lease closing backed by investment-grade credit support, and a second 400 MW Texas campus directly behind them. Diligence isn't a promise — it's a file we can show you. Qualified tenants can request the diligence index for any campus below.
Existing powered building with expansion land and fiber-rich connectivity — the fastest path to energized capacity in the portfolio.
Phase I environmental complete. Geotechnical complete. Full civil site plan issued. Substation upgrade in progress with the serving utility.
Anchor behind-the-meter campus: a turbine, reciprocating-engine, and BESS microgrid with primary firm gas transport and LNG backup.
50% civil, electrical, mechanical, and structural design sets complete. Gas quality and deliverability validated against historical data. Permitting pathway confirmed with local authorities.
Compact campus with deep expansion headroom, sized to energize fast and scale with the tenant.
A direct repeat of the Project Sky playbook: one building, one substation, one schedule measured in months.
A reference design, refined campus by campus. Built for what AI infrastructure actually requires — and for the communities that host it.
Connectivity gets the same discipline as energization. Our reference campus architecture is aligned to the hyperscaler "golden rule" standards for physical diversity — four fully diverse fiber ingress/egress routes and four fully diverse points of entry.
Dual regional rings: a northern ring reaching major metro interconnection points and multiple interstate long-haul routes; a southern ring with diverse long-haul reach of its own. True north/south/east/west multipath diversity to every major AI and cloud cluster.
Direct integration into the region's primary carrier hotel, with access to every carrier present in the facility — wavelength services across all major carriers, plus dark fiber options including a potential dedicated dark fiber loop.
The network is designed to be spliceable without impacting carrier operations — expansion and maintenance without service risk.
Single-permit route with no new easements required, fully microtrenchable for a rapid construction schedule — built on the most recent large-scale fiber infrastructure in the region, using premium modern glass.
The constraint on AI is deployed capacity. Our development system is engineered around one number: time to energization.
Land, power path, gas supply, long-lead equipment, and community agreements move in parallel — before construction mobilizes.
Substation, generation, shell, and fit-out are sequenced as one integrated program under a national delivery team, not handed off in series.
Phased delivery brings capacity online in tranches, so tenants deploy against energized megawatts while the campus builds out around them.
For grid-powered campuses of this class, where site and interconnection conditions allow. That is what the supply chain, the reference design, and the delivery system are built to do.
For an AI tenant, the schedule is the risk — every quarter in someone else's queue is deployed compute lost. Tell us your deployment profile — megawatts, density, geography, timeline — and we'll respond with options from the portfolio.
Discuss capacity