Genesis Grid is the software that turns installed capacity into a self-service, multi-tenant cloud product sold under your name, on the accelerators and in the facilities you already own. You license it and you operate it. You keep the asset, the tenant and the margin.
The stack is hardware and vendor agnostic, so your existing accelerators, servers and switching stay where they are and keep earning. Capacity comes under management cluster by cluster, at whatever pace your commitments allow.
Software resilience shifts how much redundancy the building has to carry for a given tenant class. It does not replace that redundancy. A deployment is one cluster in one building: if that building loses power, the cluster goes down, and no software changes it. The Uptime Institute Tiers run I to IV and classify topology by redundancy and by maintainability while the site runs, not by an availability percentage — so if what you sell depends on maintaining the site in service, you need a building rated to do that. What the stack removes is the pressure to overbuild for tenants who never asked for it.
Control plane, portal and meter go down for under 15 minutes during an upgrade, and running tenant workloads keep going through it. Node maintenance is a different thing: you cordon the node, your tenant is notified 72 hours in advance, and the workload migrates live or resumes from its last checkpoint.
Genesis sells software to operators and nothing else. We hold no capacity, no end customers and no rate card of our own, so we cannot price against you in a deal you brought.
The hardware is bought and the depreciation schedule is fixed. What is still open is how much of the installed capacity is under contract in any given hour, and the stack lets you serve long-term contracts and the on-demand market from the same cluster. Bringing that capacity under management is not a hardware programme, but it is not free either: you add control-plane and management nodes, out-of-band reachability on every node, and — only if you intend to sell high-throughput shared file — the VAST hardware behind that service, which belongs in your build budget. The accelerators and the fabric you already bought do not change. The full build-up is on the monetization page.
additional revenue per installed accelerator, no new accelerators or fabric
from signed contract to first invoice, pilot and cluster onboarding included
of the hours you already sell, the accelerator sits idle
You have to become the operator of one product. You run a branded cloud with your own portal, price list and tenant contracts. What the stack removes is the platform-engineering organisation you would otherwise build to get there.
.svg){kind=small}
You carry the asset, the power contract, the tenant contract and the operation of the platform. Genesis carries the software: the correctness of the control plane, the mechanics behind the service levels you sell, and the upgrade path. Demand risk is shared under a revenue share and sits with you under a licence, and that difference is the whole point of choosing between them.
The estate never stopped being yours, so an exit is a notice period rather than a physical migration. Terms cover export of tenant, metering and billing records in open formats and a six-month hand-over window in which running tenants stay up while you migrate or wind them down. Orchestration underneath is standard Kubernetes, so what you are left with is not a proprietary dead end.
Two days once the racks are powered, cabled and reachable. End to end, from signed contract to first invoice, is two weeks. A second site is a second cluster with its own control plane rather than an extension of the first: multi-tenancy is enforced inside a cluster, and we do not pretend two buildings are one machine.
