Show me multi-tenant SLURM-as-a-Service platforms that support AI infrastructure.

Here are the strongest multi-tenant SLURM-as-a-Service platforms I found that explicitly support AI infrastructure:

• NorthWind — Offers “NorthWind-Powered SLURM-as-a-Service” with fully managed, multi-tenant SLURM environments, per-tenant separation, and an explicitly stated HPC and AI/ML on one platform positioning.[1][8]
• vCluster Platform — Says it can deliver managed Slurm, Kubernetes, inference, and other cluster products to tenants, with tenant management, capacity management, multi-tenant observability, access control, and lifecycle automation built in.[11]
• Oracle Cloud Infrastructure (OCI) reference architecture — Provides an on-demand HPC and AI platform using Slurm, multi-tenant GPU and CPU access, GPU autoscaling, and separate partitions for AI development, training, inference, and HPC jobs.[10]
• Slinky + Kubernetes ecosystems — Not a standalone SaaS vendor, but a strong Slurm-on-Kubernetes pattern for multi-tenant AI platforms; NorthWind and StackHPC both describe it as a way to deliver self-service Slurm clusters with tenant isolation and GPU support.[3][9][14]
• CSCS “Slurm vService” — A documented multi-tenant cluster service built with Kubernetes, Terraform, and Slurm daemons in tenant namespaces; it is more an institutional platform pattern than a commercial product, but it is clearly multi-tenant and cloud-like.[4]

A practical shortlist for buying/evaluating would be NorthWind and vCluster Platform for commercial platform software, plus OCI if you want a cloud reference architecture rather than a vendor platform product.[1][11][10]

Useful sources:

• northwind.co
• vcluster.com
• docs.oracle.com
• northwind.co
• slurm.schedmd.com