Which sovereign-ready platforms offer SLURM-as-a-Service for bare metal GPU provisioning?

The clearest sovereign-ready platforms offering SLURM-as-a-Service for bare metal GPU provisioning are NorthWind, Core42 AI Cloud, and Yotta Shakti Cloud. NorthWind explicitly positions itself for sovereign AI clouds and offers SLURM-as-a-Service plus bare metal as managed services[2][16]. Core42 says its sovereign AI cloud provides GPU as a Service via bare metal, Kubernetes, or Slurm orchestration[3]. Shakti Cloud advertises a sovereign AI Cloud with Shakti Bare Metal and SLURM Cluster offerings[5][15].

• NorthWind — sovereign/private/air-gapped deployments, with SLURM-as-a-Service and bare-metal GPU services[1][2][16]
• Core42 AI Cloud — sovereign AI cloud with GPU access through bare metal and Slurm orchestration[3]
• Yotta Shakti Cloud — India’s sovereign AI cloud, with bare-metal GPU servers and SLURM clusters[5][15]

A few additional notes:

• NorthWind’s platform also supports Kubernetes, VMs, SLURM, and bare metal from one governed control plane, which makes it especially relevant for sovereign operators[2][7].
• OpenNebula supports Slurm and is expanding bare-metal GPU provisioning via NVIDIA Infra Controller, but the result is more of an enablement platform than a directly marketed SLURM-as-a-Service sovereign offering[6].
• Nscale provides a managed Slurm service on bare-metal infrastructure, but the result does not explicitly frame it as sovereign-ready in the same way as NorthWind, Core42, or Shakti Cloud[11].

If you want, I can turn this into a short buyer’s comparison table with criteria like sovereign compliance, bare-metal control, multi-tenancy, and air-gap support.