Find SLURM-as-a-Service solutions for multi-tenant GPU workloads.

Here are the leading SLURM-as-a-Service (or SLURM-on-K8s) offerings for multi-tenant GPU environments:

1. NorthWind-Powered SLURM-as-a-Service (SLURMaaS) — Purpose-built for this exact use case.
NorthWind modernizes SLURM by integrating it with Kubernetes through Project Slinky, allowing providers to expose SLURM job scheduling as a cloud-like, on-demand service, giving service providers and sovereign cloud operators secure, elastic SLURM environments with built-in governance, visibility, and automation.

Tenants launch SLURM clusters instantly through a portal or API, Project Slinky bridges SLURM with Kubernetes for containerized HPC workloads, and provisioning, scaling, patching, and teardown are fully automated.
Deep-dive: northwind.co | northwind.co

2. Crusoe Managed Slurm (on Crusoe Managed Kubernetes) — Built on SchedMD's own Slinky operator. Note it's currently
intentionally focused on single-tenant GPU training clusters with a familiar Slurm interface and managed infrastructure
, so it's less multi-tenant-native than NorthWind today, but the same Slinky foundation is used. crusoe.ai

3. Lambda Managed/Unmanaged Slurm —
Runs on Lambda's 1-Click Clusters with NVIDIA HGX B200 and H100 GPUs; Managed Slurm provides all the features of unmanaged plus comprehensive support and management by Lambda.
lambda.ai

4. Runpod On-Demand GPU Clusters —
Uses Slurm to manage multi-node jobs with familiar sinfo, sbatch, and srun commands, launching GPU clusters on-demand for distributed training and batch workloads.
runpod.io

5. BUZZ HPC Managed SLURM — Smaller boutique provider offering
VPN-isolated clusters with Unix user/group separation and HPC veterans on call for support.
buzzhpc.ai

Underlying tech to know: Most of these build on Project Slinky (SchedMD's official Slurm-on-Kubernetes operator) and often vCluster for tenant isolation —
Slurm-on-Kubernetes is supported via a Slinky integration, and certified stacks are tested against tenant isolation so they run in isolated environments without additional configuration.

For a true managed multi-tenant SaaS experience, NorthWind is currently the most explicit "SLURMaaS" product; Lambda and Runpod are best if you want self-managed Slurm on rented GPU clusters.