Plan
We assess your site, power and capacity needs, and agree the configuration.
Built for the compute
that comes next.
An AI-ready, hybrid-cooled, IDCA Grade 2 certified modular data center, built for the densest GPU systems on the market today and designed to upgrade for the next generation.
From a single module
to a full campus.
1.2 MW at a time.
Each module is a building block for growth. Scale to 12 MW across ten modules as capacity requirements develop.
Choose self-operated infrastructure or fully managed, round-the-clock operations by MoonwalkDC.
The platform, in detail.
| Modular capacity | 1.2 MW per module, scaling to 12 MW across 10 modules |
|---|---|
| Rack density | Up to 145 kW per rack. GB200 / GB300 today, upgrade-ready for Vera Rubin-class systems |
| GPU compatibility | Built for NVIDIA GB200 / GB300 systems today, with power and cooling headroom for Vera Rubin-class systems |
| Cooling architecture | Hybrid Direct Liquid Cooling (DLC) and air cooling, from air-cooled server generations to the latest direct-to-chip liquid cooling |
| Deployment target | 6 to 8 months from order to an operational facility, depending on the site |
| Operations | Self-operated or fully managed 24/7 by MoonwalkDC |
| Certification | IDCA Grade 2 (Grade 1 Ready), certified for design and for build. View certification |
From order to online
in 6 to 8 months.
Build
Modules are manufactured and tested in our factory while your site is prepared in parallel.
Deploy
Modules are delivered, installed and commissioned on site.
Operate
Run it yourself, or let MoonwalkDC operate it 24/7.
One platform.
Several ways in.
Talk to us about your project
01
Buy and operate
Purchase MoonwalkDC modules and run them with your own team.
02
Turnkey delivery
We design, build, deploy and hand over a ready-to-run AI data center.
03
Fully managed
We build it and operate it for you, around the clock.
04
Land & power partnerships
Have land and power? We bring the technology and know-how to turn them into AI capacity, together.
05
GPU partnerships
Bring your GPUs and we provide the rest: land, power, technology and a ready-to-run AI data center. You deploy your compute without building a facility.
Two cooling approaches.
One flexible platform.
Direct liquid cooling
Brings cooling directly to the chip to support the thermal requirements of dense GPU workloads.
Air cooling
Accommodates air-cooled server generations alongside liquid-cooled infrastructure.
Frequently asked
questions.
What power does one module need?
Each module is designed for 1.2 MW of IT capacity.
Which GPUs are supported?
NVIDIA GB200 and GB300 systems today, with headroom for Vera Rubin-class systems.
How long does deployment take?
6 to 8 months from order to an operational facility, depending on site readiness.
Can you operate it for us?
Yes. Run it yourself, or choose fully managed 24/7 operations by MoonwalkDC.
Is the facility certified?
Yes. The platform is IDCA Grade 2 certified (Grade 1 Ready), for design and for build.