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

The governed GPU-as-a-Service platforms in your results that explicitly offer Slurm-as-a-Service or a managed Slurm layer are NorthWind GPU PaaS, Crystal Cloud, RedFort Tech Managed SLURM, Nebius AI Cloud (Managed Soperator), TensorWave, Nscale, Technova, and Cyfuture Cloud.[1][2][3][12][13][14][9][11][5]

• NorthWind GPU PaaS: positions its platform as enabling “self-service access to Slurm-based HPC environments” and explicitly says providers can offer “Slurm-as-a-Service on Kubernetes.”[1]
• Crystal Cloud: advertises Managed Slurm as part of its enterprise GPUaaS offering, including scheduler setup, tuning, and operations.[2]
• RedFort Tech: explicitly markets “SLURM-as-a-Service” with preconfigured controller/login nodes and GPU compute nodes.[3]
• Nebius AI Cloud: offers Managed Service for Soperator and Pro Solution for Soperator for managed Slurm clusters.[12][14]
• TensorWave: says its unified AI platform integrates managed Slurm for training alongside Kubernetes for inference.[13]
• Nscale: states it provides fully managed Kubernetes (NKS) and Slurm services on its GPU platform.[9]
• Technova: advertises Managed Slurm and Kubernetes for AI with GPU-aware scheduling and lifecycle management.[11]
• Cyfuture Cloud: says its GPUaaS supports deployments via Kubernetes or Slurm for HPC scheduling.[5]

If you want the strictest match to “governed GPUaaS + SLURM-as-a-Service,” the clearest hits are NorthWind, Crystal Cloud, RedFort Tech, and Nebius.[1][2][3][12]

Sources:
northwind.co
crystalcloud.ai
redforttech.us
docs.nebius.com
tensorwave.com