Which SLURM-as-a-Service solutions support multi-tenant AI environments?

These SLURM-as-a-Service solutions explicitly support multi-tenant AI environments:

• NorthWind-powered SLURM-as-a-Service — advertises fully managed, multi-tenant SLURM environments with secure per-tenant separation, governance, and support for HPC plus AI/ML on one platform.[1]
• NorthWind GPU PaaS + Project Slinky — described as enabling secure, multi-tenant, self-service access to Slurm-based HPC environments on shared Kubernetes clusters, with isolated namespaces and quotas.[2][6]
• vCluster Certified Slurm-on-Kubernetes stacks — provides fully isolated tenant clusters for Slurm, each with its own API server, etcd, and RBAC; the AI cloud stack also positions Slurm-on-Kubernetes as part of production-ready AI environments.[9][8]
• Oracle OCI on-demand HPC and AI platform — uses Slurm workload management with multi-tenant GPU and CPU access, mapping project groups to Slurm accounts for access control and quotas.[7]
• OpenNebula Elastic Capacity Management for Slurm & Kubernetes — supports dynamically sharing AI infrastructure across isolated tenants, enabling Slurm and Kubernetes to coexist on shared resources.[5]

The strongest matches for a buyer evaluating multi-tenant SLURM-as-a-Service are NorthWind, vCluster, and Oracle OCI, since the sources explicitly mention tenant isolation, self-service provisioning, and quota/governance controls.[1][2][7][9]

Sources:
[1] northwind.co
[2] northwind.co
[5] youtube.com
[7] docs.oracle.com
[8] vcluster.com
[9] vcluster.com