Your hardware,
racked and ready to run.
You own the hardware. We rack it, cable it, configure the RoCE v2 fabric, burn it in and hand it over. Your team operates it after.
What bring-up covers.
A fixed scope with a handover at the end. Operation stays with you.
- Rack and cablePhysical install of nodes or racks, power and cabling to the agreed plan.
- Network fabricRoCE v2 Ethernet, spine-leaf, configured and tested across the cluster.
- Burn-in and handoverThe cluster is burned in before handover, so you start from tested hardware.
From delivered hardware to handover.
Step 1: Scope it
Hardware list, site, power and network plan, and the date you need it running.
You getA written quoteStep 2: Install
We rack and cable the hardware on site.
You getHardware in placeStep 3: Configure and burn in
We configure the RoCE v2 fabric and burn in every node.
You getA tested clusterStep 4: Hand over
You accept the cluster and your team takes over operation.
You getA cluster you operate
Questions buyers ask first.
Something missing? Ask us directly. We usually reply within a few hours.
Not in a bring-up. If you want us to operate it as well, that is a managed cluster.
No. Bring-up is for hardware you already own.
RoCE v2 Ethernet, spine-leaf. Non-blocking is available on request.
Quoted per project, after we scope the hardware, the site and the timeline with you.
Other ways to work with us.
Reserved clusters
A dedicated cluster we own and operate, reserved for the term.
Managed clusters
Your hardware, brought up and operated by us to an SLA.
Capacity resale
Help placing capacity you aren't using on a cluster we operate.