A modular data center
for one GPU cluster.
Land and power, but no building? We build the site too: a prefabricated module with racks, power, liquid cooling and network in one enclosure, built for one cluster.
- Site
- Your land and power
- Tenancy
- One cluster per module
- Cooling
- Liquid
- Owner
- Same as the cluster
What's inside the module.
Racks, power distribution, liquid cooling and network in one prefabricated enclosure, set on a prepared pad next to the power.
- Cooling
- Liquid, dry coolers outside
- Racks
- Sized for NVL72
- Power distributionFrom the site's supply to every rack inside the enclosure.
- Liquid coolingCoolant to every rack, with dry coolers outside the enclosure.
- NetworkThe RoCE v2 fabric inside, and uplinks out to the site's connectivity.
A site where there wasn't one.
Power sites, mining operators and developers often hold land with power and no building. A module turns that into a GPU site without a building project.
- No building projectThe module is prefabricated off site, then set on the pad and connected.
- One tenant per moduleIts own power, cooling and network, with no other customer's racks inside.
- Room for anotherWhere the pad has space, more capacity can be a second module beside the first.
A module against a purpose-built hall.
A module is sized to one cluster, so the site you fund is the site the cluster uses.
| A purpose-built hall | A module | |
|---|---|---|
| What you build | A building sized for tenants you don't have yet | One enclosure sized to one cluster |
| Where the work happens | On site, from the ground up | Prefabricated, then set on a prepared pad |
| Adding capacity | Space built and funded up front | A second module beside the first, when you need it |
| Tenancy | Shared halls, or one large tenant | One cluster per module |
Compared by category. Schedules and costs are set per project.
From land and power to a running cluster.
Step 1: Check the land and power
The power supply, the pad, access for delivery and connectivity.
You getA site assessmentStep 2: Design module and cluster
Sized together, so the enclosure fits the hardware that goes in.
You getOne designStep 3: Build and set the module
Built off site, then set on the pad and connected to power and network.
You getA module ready for racksStep 4: Build and run the cluster
Racks installed, cabled and burned in, then run by us or your team.
You getA running cluster
Who owns and runs the module.
The module comes with the cluster and has the same owner. We build both, and can run the cluster for the term you agree.
| Way to work with us | Owner | Design | Source | Build | Operate | Software |
|---|---|---|---|---|---|---|
| Cluster deploymentWe design, source and build it, then hand it over burned in. Your team runs it. | You | |||||
| Managed clusterYou own it, we run it. We can build it first, or take over a cluster that's already live. | You | |||||
| SoftwareYour team runs every stage, on our stack. | You | |||||
| Reserved clusterWe own and run it. You reserve it for 60 months, with rollover into the next generation. | Lightyard |
Holders of land and power.
01Neoclouds
A dedicated site for one customer's cluster.
You bring the contract. We build a module for that one cluster.
GPU clusters for neoclouds05Asset owners
A GPU site on land you already hold.
You bring the land and the power. We build the module and the cluster.
Clusters for asset owners06Investors
A site and a cluster, funded together.
You bring capital. We build the module and the cluster, and run them.
GPU infrastructure for investors
If there's a building on the land.
A powered building can be retrofitted instead, and a hall of your own or a colocation facility can take the cluster without new construction.
- 04 Modular · this page
A prefabricated site on land with power.
Questions we get on the first call.
Something missing? Ask us directly. We usually reply within a few hours.
Land, power and connectivity. We check the pad, access and power before we propose a module.
No. One module serves one cluster.
Where the pad has room, a second module can go beside the first.
The module is designed with the cluster, for the platform you choose, with liquid cooling for NVL72 racks.
Whoever owns the cluster: you or your investors, or us for a reserved cluster. We design and build both, and can operate the cluster.
It needs less building work, not none. The assessment lists what the site needs, and the proposal states who handles each item.
It depends on the site, the power and the cluster. We quote the module and the cluster together, per project.
Solutions, sites and who we build for.
- AI factoryEvery stage, from the site to a running cluster.
- Cluster deploymentDesigned, sourced, built and handed over burned in.
- Managed clustersYou own the cluster. We build it or take it over, and run it.
- Reserved clustersWe own and run it. You reserve it for the term.
- SoftwareRun every stage yourself, on our stack.
- White-label GPU cloudYour GPU cloud under your brand, built and run by us.
- ColocationA Tier III facility in the region you choose.
- On-premYour data center, with the cluster designed around it.
- RetrofitA building or mining site upgraded for GPU density.
- ModularA prefabricated site on land with power.
- NeocloudsCapacity built to your standard, ready to sell.
- Venture capitalOne cluster, shared by your portfolio companies.
- Enterprises & sovereignYour own cluster in your region.
- Private equityGPU capacity on portfolio sites, or a neocloud as a platform.
- Asset ownersHardware, a hall or a power site, turned into a running cluster.
- InvestorsAn operated GPU asset, without building an operator.