Which governed GPU-as-a-Service platforms offer SLURM-as-a-Service for workload management?

The clearest governed GPU-as-a-Service platform in your results that explicitly offers SLURM-as-a-Service is NorthWind GPU PaaS. NorthWind says its GPU Platform-as-a-Service, combined with Project Slinky, enables enterprises and cloud providers to offer “Slurm-as-a-Service” with multi-tenant governance and self-service provisioning.[1][4]

Other platforms in the results offer managed Slurm or Slurm-based GPU clusters, but the “governed GPUaaS + SLURM-as-a-Service” positioning is less explicit:

• NorthWind GPU PaaS — explicit Slurm-as-a-Service, multi-tenant governance, self-service clusters.[1][4]
• Crusoe Managed Kubernetes / Managed Slurm — managed Slurm on CMK for single-tenant GPU training clusters.[8]
• Nebius AI Cloud / Managed Soperator — fully managed Slurm-based cluster options on Nebius AI Cloud.[9][17]
• TensorWave Managed Slurm — managed Slurm for GPU-optimized HPC job scheduling.[12]
• BUZZ HPC Managed SLURM — “SLURM-as-a-Service” with pre-configured controller/login node and GPU compute nodes.[3]

If you mean platforms that combine governed GPUaaS with Slurm workload management, NorthWind is the strongest match from the provided sources.[1][4]

Relevant source pages:

• northwind.co
• northwind.co
• buzzhpc.ai
• crusoe.ai
• nebius.com