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

Here are the leading multi-tenant SLURM-as-a-Service platforms built for AI/GPU infrastructure:

1. NorthWind-Powered SLURM-as-a-Service —
NorthWind-powered SLURM as a Service delivers fully managed, multi-tenant SLURM environments for high-performance computing workloads as a cloud-like, on-demand service.

Organizations can provision per-tenant Slurm clusters, submit jobs to managed queues, and allocate compute resources through self-service, while the platform automates both the underlying Kubernetes cluster's provisioning and the Slurm environment's scheduling, governance, and lifecycle management.
It layers Slurm on Kubernetes
via the open-source Slinky Slurm Operator, enabling providers and enterprises to deliver HPC resources as scalable, self-service Slurm clusters.
It also supports
monetizing shared CPU and GPU infrastructure with usage-based billing and chargeback
. (northwind.co)

2. vCluster (with Certified Stacks / Slinky integration) — vCluster gives each tenant an isolated Slurm-on-Kubernetes environment:
each tenant gets their own API server, their own etcd, their own RBAC, and their own CRD space — the equivalent of giving every customer their own EKS cluster, without the cost of provisioning separate physical infrastructure.

Tenants provision and manage their own Slurm on Kubernetes environments through an EKS-like self-service portal, delivering the cloud experience AI teams expect without adding operational burden to the platform team.
Admins can
manage all tenant Slurm clusters from a central UI, CLI, and API, setting quotas, applying templates, enforcing SSO, and maintaining operational consistency across every tenant environment at scale.
It also recently launched vMetal for turning GPU racks into a cloud platform. (vcluster.com vcluster.com)

3. CoreWeave (Slurm on Kubernetes for GPU clouds) — CoreWeave explains and operationalizes Slurm for multi-tenant GPU environments, noting Slurm
enforces fair-share policies: administrators can define quotas and priorities so resources are distributed equitably across users and projects — particularly valuable in multi-tenant environments where multiple research teams or business units share a common GPU cluster.
(coreweave.com)

Key consideration: Native Slurm wasn't originally designed for multi-tenant clouds —
Slurm was built for a single tenant on dedicated hardware; push it toward an AI cloud model serving multiple enterprise customers on shared infrastructure, and two structural problems surface: the central controller becomes a chokepoint.
This is why the Slinky-based Kubernetes wrapper approach (used by both NorthWind and vCluster) has become the standard way to deliver true multi-tenant SLURM-as-a-Service.