Prefabricated, AI-Ready Container Datacenters — RDHx-First, Lower Capex
Modular HPC and AI capacity without a construction division. We design and integrate container-module datacenters and deliver them through vetted EPC partners.
Air-First, Growth by Module: The Density Ladder
Not every rack needs to be 150kW. The market's binary "air vs. liquid" debate ignores the pragmatic middle — RDHx-cooled modular pods handle AI and HPC densities at a fraction of direct-to-chip capex, and when workloads demand more, you scale by swapping or adding factory-built containers — not by retrofitting a building.
How We Deliver
Requirements & Sizing
IT load, density ladder mix, growth path, site constraints — capacity engineered to the workload, not the vendor catalog.
Solution Architecture
Container-module layout, RDHx-first cooling design, power distribution, and network topology — as a tender-ready design package.
Vendor & EPC Tender
Container-module vendor selection and EPC tender, run with the same rigor we apply to our own four-region program.
Delivery Management
Milestones, hold-points, and supervision of your EPC — factory witness tests, site works, and schedule defended on your behalf.
Commissioning L1–L5
Factory acceptance through integrated systems testing — the building proves itself before you accept it.
Handover to Operations
As-builts, documentation, and a direct path into our Managed Operations — the same team that delivered it runs it.
RDHx-First vs. D2C-Everywhere
| D2C-Everywhere Design | QoreLayer RDHx-First | |
|---|---|---|
| Cooling capex | Highest — CDUs, manifolds, liquid loops everywhere | Materially lower — liquid at the rack rear, not the chip |
| Deployment speed | Slower — liquid commissioning across all capacity | 12–18 months to live, phased per module — vs 18–36 months ground-up |
| Complexity & skills | Liquid operations expertise required day one | Standard operations; D2C skills only when added |
| Future flexibility | Locked into one cooling doctrine | Swap or add containers — factory-integrated upgrade path, no facility retrofit |
| Where it wins | Uniform ultra-density estates | Mixed AI/HPC + enterprise — most real workloads |
Reference Module — 500kW Class
Our standard delivery unit, built with our manufacturing partner to our specification — one container, factory-integrated and tested.
| IT Load | 500 kW |
| Racks / Density | 10 racks · 50 kW per rack |
| Tier Level | T3-equivalent (concurrently maintainable) |
| Cooling | RDHx (10 × 65 kW) · air-cooled chiller with free cooling |
| Power | 2 × 600 kVA modular UPS (2N) · 5 min autonomy |
| DCIM | Integrated — full sensors, Modbus/SNMP |
| Fire Suppression | Clean-agent (optional) |
| Footprint | Single 12.8 × 3.0 × 7.3 m container |
Densities scale up and down the ladder: single-rack micro modules (25–90 kW) for edge and pilots, D2C containers (80 kW/rack) and immersion (100 kW/rack) where workloads demand it — added as containers, so the site grows without new construction.
Pricing
| Line | Price | Notes |
|---|---|---|
| Reference 500kW module (10 racks, T3) | from ~$3.5M | ≈ $7M/MW equivalent, AI/HPC-ready, RDHx-cooled |
| Delivery management | 4–8% of CAPEX | Tender, supervision, commissioning oversight |
| Owner's engineer add-on | 3–5% of CAPEX | Full lifecycle representation, contractually separated |
| Capex gainshare | 15–25% of verified savings | Vs. agreed baseline design, capped both ways |
Modular deployments typically land at $7–9M/MW versus $10–12M/MW for traditional Tier III builds and $15–20M+/MW for AI-optimized ground-up. The deeper savings come from phasing — 12–24% lower 10-year TCO when you build only what you need now. Prices last reviewed September 2026.
What Moves Your Price
IT Load
Total capacity in kW and module count.
Density Mix
Ratio of air / RDHx / D2C pods in the ladder.
Site Readiness
Grid connection, civil works, permits.
Redundancy & Timeline
N+1 vs 2N, and how fast you need it live.
FAQ
RDHx or direct-to-chip — which do I need?
RDHx handles mid-to-high rack densities at a fraction of D2C cooling capex; D2C becomes necessary for the highest-density GPU pods (typically beyond 80kW/rack; immersion from 100kW/rack). We design the ladder: air for general compute, RDHx for AI/HPC pods, D2C only where truly required — and when densities rise you upgrade by swapping the container for a D2C or immersion module, or adding one, at factory cost.
Do you build the datacenter yourselves?
No — by design. We architect the solution, tender the modules and EPC scope, and manage delivery through vetted construction partners. You get factory-quality modules and accountable delivery without owning a construction division.
How fast can it be live?
Typically 12–18 months from order to live capacity — supplier lead times for high-density power and cooling components are the binding constraint, not construction. That still beats the 18–36 months of ground-up builds, and capacity arrives in phases: each module powers on as it lands.
How does modular capex compare to traditional?
Modular typically lands at $7–9M/MW versus $10–12M/MW traditional Tier III and $15–20M+/MW for AI-optimized ground-up builds. Phasing delivers the deeper win: 12–24% lower 10-year TCO when capacity tracks demand. Our capex gainshare shares the verified savings with you — see how success fees work.