Working with partners
Some of our work comes through partners: hardware resellers, system integrators, and GPU cloud or colo providers whose customers need the software stack brought up on the hardware they've sold. On those deals we deliver the layer between the metal and the workloads - provisioning, network fabric, Kubernetes or Slurm, scheduling, monitoring, and handover to the customer's team.
We're flexible on how engagements are structured and branded, and the usual starting point is a single deal rather than a formal agreement.
If you have a customer or a deal where this fits, get in touch.
Our engineers contribute upstream to the NVIDIA stack - the work is public.KAI Scheduler (NVIDIA)NVIDIA Network OperatorMellanox ipoib-cniNVIDIA Certified EngineersApache Software Foundation member