Solomon Data Systems develops right-sized AI data center campuses across America — 20 to 400 MW sites with secured power, behind-the-meter generation, and a supply chain that delivers in months, not years.
Figures as of June 2026. MW figures are critical IT capacity in active development; lease capacity in commercial closing is subject to definitive agreements; pipeline figures indicative. † Delivery timeline for grid-powered campuses of this class, where site and interconnection conditions allow.
Right-sized moves faster. Our first three campuses — with 240 MW in commercial lease closing, backed by investment-grade credit support — are built to energize quickly, prove the model, and repeat it.
Existing powered building with expansion land and fiber-rich connectivity. The fastest path to energized capacity in the portfolio.
Anchor behind-the-meter campus: a turbine, reciprocating-engine, and BESS microgrid with primary firm gas transport and LNG backup — power that never waits in a queue.
Compact campus with deep expansion headroom — a direct repeat of the Project Sky playbook, sized to energize fast and scale with the tenant.
For grid-powered campuses of this class, where site and interconnection conditions allow. Long-lead equipment managed from day one; design, permitting, and procurement run in parallel.
Data center capacity fails for predictable reasons: power that never arrives, land that never permits, equipment that never ships, communities that never consent. We secure all four before the first rack arrives.
Grid interconnection where it works, behind-the-meter microgrids where it doesn't. Once the tenant signs, our capital builds the substations, feeders, and protection upgrades — utilities and their members never subsidize our load.
Controlled, entitled land ready for dense AI loads — qualified by our proprietary siting and diligence technology before we ask for a tenant's time.
Generation equipment on 8-month lead times against industry quotes of 24–36; high voltage transformers and breakers in 6–12. The supply chain is the schedule.
We covenant what others talk about: we sign binding commitments on water, light, sound, and emissions with host counties and serving utilities.
Full campuses engineered for single anchor tenants running dense AI workloads — not retail colocation retrofitted for the moment.
Liquid-cooled, DC-power-ready designs that deliver 300 kW+ per rack now — for the inference and training fleets actually shipping, not the ones being announced.
Water-free heat rejection tuned to a 1.1–1.2 annualized PUE target — less energy overhead, less strain on the grid, for the same compute output.
Harmonic management, supercapacitor support, and blended peaking-plus-baseload generation — power quality engineered into the campus, not bolted on.
Powered campuses in Georgia, Texas, and Ohio — with a 1.5 GW pipeline behind them — delivered on timelines the grid alone can't offer.
Real demand, institutional capital relationships, and covenants in writing — before we request a single study.
AI inference silicon, network fabric, long-lead electrical, plus colocation and GPU capacity through the platform.
Every Solomon campus starts where most projects stall: secured energy. Tell us what you need to deploy, and we'll show you how fast we can stand it up.
Start the conversation