A GPU cloud under your brand,
built and run by us.
Launch a neocloud without building the team behind it. We design, source, build and operate the cluster, and our software runs your console, tenants and billing under your name. You sell the capacity and keep the customers.
- Brand
- Yours
- Cluster owner
- You or your investors
- Operator
- Lightyard, to an SLA
- Customers
- Yours
What your customers see carries your name.
Your customers deal with your brand and your team. Everything behind it, from the sites to the console they log in to, is ours to build and run.
- Your brand and domainThe console, the docs and the invoices carry your name and sit on your domain.
- Your customer relationshipsYou set the prices, sign the contracts and decide who gets capacity.
- Tenants and billing includedAllocations, contracts and billing run on the same inventory as the hardware.
What a new GPU cloud needs, and who provides it.
Most of the work of starting a GPU cloud happens before the first customer signs. You can hire for all of it, or contract it and hire for the commercial side.
| Built in house | With us behind you | |
|---|---|---|
| Site and design | A design team hired before the first cluster | Our engineers, with a site found if you have none |
| Hardware | Supplier relationships built from scratch | Bought through our relationships with NVIDIA and server manufacturers |
| Build | An integrator, then acceptance tests you write | Our build team, with every node burned in before you sell it |
| Operations | A hardware operations team, hired and trained | Ours, to an SLA in the contract |
| Tenant software | A console, metering and billing, built or bought | Our cloud layer, under your brand |
| Sales and customers | Your team | Your team |
Compared by structure. The commercial side stays yours either way.
The costs are on paper before you commit.
The build and the running cost are set in the proposal. What you sell, and how much, is yours.
- Your assetTitle passes to you, or your vehicle, at delivery.
- Costs in the proposalBuild, operations and software, priced before you sign.
- Your upside, your riskPrices and utilization are yours. We don't promise either.
Summary of principal terms
White-label GPU cloud · Illustrative
- Owner
- [Your company] or its vehicle
- Title
- Passes to the owner at delivery
- Build
- Priced in the proposal, through acceptance
- Operations
- Priced in the proposal for the term, to an SLA
- Software
- Console, tenants and billing under your brand
- Prices and customers
- Set by you
- Utilization
- Yours. Not guaranteed by Lightyard
What we do and what stays with you.
The proposal sets the split line by line. The commercial side of the cloud is always yours.
- 01We build
- Site selection
- Design and bill of materials
- Hardware procurement
- Racking, cabling and fabric
- Burn-in and acceptance
You get A cluster ready to sell
- 02We run
- Monitoring and health
- Failed-part replacement
- Firmware and spares
- The console, tenants and billing
- Reports to your team
You get A cloud run to the SLA
- 03You run
- Your brand and pricing
- What you sell
- Sales and customer contracts
- First-line support for your customers
You get The customer relationship
From a plan to a GPU cloud you can sell.
Step 1: Define the cloud
Platform, size, region, and what you plan to sell: dedicated clusters, nodes or a platform on top.
You getA written proposalStep 2: Site and design
We find the site if you need one and design the cluster around your first customers.
You getA design and a dated planStep 3: Build and brand
We build and burn in the cluster while your console, docs and invoices are set up under your brand.
You getAn accepted cluster under your nameStep 4: Open for customers
You sell the capacity. We operate the cluster and the software behind your cloud, to an SLA.
You getA running GPU cloud
A managed cluster, sold under your brand.
In the ways table this is a managed cluster: you own it and we run it. White-label adds the cloud layer under your name.
| 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 |
Who launches a GPU cloud this way.
01Neoclouds
A neocloud from the first cluster.
You bring the plan and the first customers. We build the cloud behind them.
GPU clusters for neoclouds02Venture capital
A GPU cloud for your portfolio.
You bring the companies. They log in to a console under the fund's name.
GPU clusters for venture portfolios03Enterprises & sovereign
An internal GPU cloud for your teams.
You bring the teams and the region. They get a console under your name.
Private and sovereign AI04Private equity
A neocloud as a platform company.
You bring capital and a management team. We build and run what they sell.
AI infrastructure for private equity05Asset owners
Sell the capacity of a site you own.
You bring a hall, power or hardware. We turn it into a GPU cloud under your name.
Clusters for asset owners06Investors
A GPU cloud as an investment.
You bring capital and an operating partner. We build and run what they sell.
GPU infrastructure for investors
Where your cloud can run.
A Tier III facility in the region your customers need, your own site, a retrofit or a modular data center.
Questions we get on the first call.
Something missing? Ask us directly. We usually reply within a few hours.
No. The console, the docs and the invoices carry your brand, and your customers deal with your team.
Your team, as their first line. We sit behind you on the cluster and the software, under the SLA in our contract with you.
Dedicated clusters, nodes allocated to tenants, and the platform layer on top: endpoints, jobs and containers. The mix is set at the design stage.
No. Sales are yours. We build and run the cloud and don't promise utilization, tenants or revenue.
Yes. We can operate the cluster under your own console or a partner's and keep our stack to the fleet layer.
Yes. The cluster is yours, and the software can be licensed for your own team to run it on.
Most clusters are 32 to 128 nodes; larger ones are scoped case by case. Later clusters join the same cloud as you grow.
Per project: the build quoted on its own, then operations and the software 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.