An operated GPU cluster,
owned by you or your fund.
You bring the capital. We design, source, build and operate the cluster as one counterparty, and the asset stays with you or your vehicle.
- Owner
- You or your fund
- Design and build
- Lightyard
- Operate
- Lightyard, to an SLA
- Counterparty
- One
One counterparty, from design to operations.
If you've financed energy infrastructure, this is EPC and O&M, applied to GPU clusters: one team engineers, procures and builds the cluster, then operates and maintains it.
| Several contractors | One counterparty | |
|---|---|---|
| Engineering | A consultant or design firm | Our engineers |
| Procurement | Purchase orders spread across suppliers | Through our relationships with NVIDIA and server manufacturers |
| Construction | An integrator | Our build team, to our design |
| Operations and maintenance | An operator hired after handover | Us, to an SLA in the contract |
| A fault between stages | A dispute between contractors | One party answers for it |
Compared by structure, not by firm.
Who holds what.
The cluster belongs to you or the vehicle you set up. We hold the contracts to build it and to run it, not the asset.
- Title stays with youHardware is bought for the project and owned by you or your vehicle.
- Two scopes of workA build scope that ends at acceptance, then operations under an SLA.
- The capacity is yours to placeWho uses it, and on what terms, stays your decision.
What the contract covers, and what it doesn't.
The proposal sets the scope line by line. The last row is what we don't promise to anyone.
- 01Build
- Site selection
- Design and bill of materials
- Hardware procurement
- Racking and cabling
- Burn-in
- Acceptance per node
You get An accepted cluster
- 02Operate
- Monitoring
- Failed-part replacement
- Firmware
- Spares
- Reporting
- Our software
You get A cluster run to the SLA
- 03Not promised
- Utilization
- Revenue or returns
- Customers for the capacity
Most investors choose a managed cluster.
You own it and we do the rest. If you'd rather not hold hardware at all, a reserved cluster leaves it with us.
| 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 |
Where the asset can sit.
The design stage chooses the site with you: a colocation facility, a site you or a portfolio company already hold, or new capacity on land with power.
What the fund would own.
Current NVIDIA Blackwell systems, from HGX B300 nodes to GB300 NVL72 racks, with Vera Rubin on the waitlist.
- Rack scaleNVIDIA GB300 NVL72
- Next generationNVIDIA Vera Rubin
Questions we get on the first call.
Something missing? Ask us directly. We usually reply within a few hours.
Yes, the fund or a vehicle it sets up. We design, build and operate under contract and don't take title.
No. We build and operate the cluster to an SLA. Who uses the capacity, and what it earns, depends on the contracts you or your partners sign.
In structure, yes: engineering, procurement and construction, then operations and maintenance, from one counterparty. The asset is a GPU cluster instead of a plant.
A design and bill of materials before hardware is ordered, a dated delivery plan, and acceptance results for each node at handover.
It stays yours and keeps running under the operations contract. Adding or moving to a newer generation is a new scope, quoted on its own.
Yes. A data hall, a power site or land with power can host it, with a retrofit or a modular data center where needed. Without a site, we find one.
Both are in the proposal: the build as a project, operations as a term contract with an SLA.
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.