The product
The C40.
One module, one architecture: a factory-built, liquid-cooled AI data centre in a 40-foot container, run by the MASH platform. All figures below are design targets, to be validated on our prototype.
The C40 at a glance.
| Specification | C40 |
|---|---|
| Enclosure | ISO 40-ft high-cube container |
| Facility power | 800 kW |
| Cooling | Cold-plate liquid cooling, CDU and dry coolers, plus cold-climate free cooling |
| Design PUE | ≤ 1.13 (target, to be measured on our prototype) |
| Deployment | 8–12 weeks, factory to commissioned (design target) |
| GPUs | GPU-agnostic |
Design features
Built into every C40.
Cooling
Cold-plate liquid cooling with a CDU and dry coolers, plus cold-climate free cooling. Design PUE target ≤ 1.13, to be measured on our prototype.
Network
Spine-Leaf fabric with isolated out-of-band management.
GPU-agnostic
Works with the GPU servers you choose, selected per deployment.
Scale
Add C40 modules side by side to grow capacity without redesigning the site.
The MASH platform.
Unified management, orchestration, and operations across every base: DCIM, BMS and EMS, multi-tenancy, alerting, billing, and 24/7 remote operations.
Three ways to buy
Buy it, finance it, or lease-to-own it.
Buy the C40
Purchase C40 modules outright and site them on your ground.
Finance or lease-to-own
Equipment financing or lease-to-own, with optional managed operations by MASH.
Your choice of GPUs
We integrate the GPU servers you choose. Compute services are on our roadmap.
Want the full spec sheet?
We’ll send the C40 design specifications and walk you through what a deployment would take.