Request access to the Spinifex sandbox. Request access
Layer_1
Deployments

Three deployment shapes. One platform.

Pick the deployment that matches your reality: partner Neocloud, your own data centre, or the edge. The AWS-compatible surface on top is the same.

01 Neo-clouds

Lift and shift onto a partner Neocloud.

Neocloud and GPU specialists have the iron and the prices. What they usually lack is the AWS-compatible surface that lets enterprise workloads land unchanged. Spinifex is that surface.

What you get

EC2 / EBS / S3 / VPC on a Neocloud's fleet. Workloads built for AWS run unchanged. Change an endpoint, keep the software.

Why it works

A migration that used to be a multi-year re-platform becomes a config change. The economics close immediately.

Best fit

GPU-heavy AI workloads, repatriation projects, organisations that want to leave AWS but don't want to own the metal.

For Neocloud operators

Drop Spinifex onto your platform to capture AWS repatriation demand without a custom integration per customer.

GPUH100 / H200 / B200tenancyrepatriationpartner network
02 On-premise

Bring your own hardware. Host Spinifex in your DC.

Your data centre, your network, your governance. Spinifex runs as a real distributed cloud across whatever hardware you bring, and it exposes the AWS APIs your software already speaks.

What you get

Multi-node HA cluster integrated with your storage and networking. AWS-compatible APIs, AGPL core. You own the stack end to end.

Why it works

A predictable bill (no egress surprises), data in the jurisdiction you choose, auditable open-source platform.

Best fit

Regulated workloads, sovereign data, organisations with existing DC capacity and a mandate to keep workloads on-prem.

Reference architectures

Single-node, 3-node HA, multi-rack. Storage via Predastore or bring your own. Reference configs and install guides in docs.

data centremulti-nodeHAAGPLsovereignauditable
03 Edge

Cloud where the cloud can't reach.

Sites that go offline. Vehicles, vessels, drilling rigs, remote clinics, factory floors, forward operating bases. Anywhere the link to the hyperscaler is intermittent, expensive, or simply not there.

What you get

A real AWS-compatible cloud on hardware at the site. Survives full disconnection. Same APIs your software already uses.

Why it works

Compute next to the data. Decisions that have to be instant happen locally, with no round trip to a distant region.

Best fit

Defence, energy, mining, maritime, manufacturing, healthcare, automotive. Anywhere the work doesn't happen in a data centre.

Form factors

Single-box appliance, rugged rack, mobile shelter, ship-borne cabinet. Spinifex runs on whatever you can power.

DDILoffline-firstapplianceruggedmobilelow-power
Whichever shape

Try Spinifex against your own workload.