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

SLURM-as-a-Service Solutions for Multi-Tenant GPU Workloads

1. NorthWind-Powered SLURM-as-a-Service — This is the closest to a purpose-built "SLURMaaS" offering.
It delivers secure, fully managed SLURM clusters as elastic, multi-tenant HPC services with lifecycle automation and governance using the NorthWind Platform.
It combines Project Slinky (SchedMD's Slurm-on-Kubernetes operator) with NorthWind's GPU PaaS:
each user operates in their own NorthWind-managed Kubernetes namespace for strong tenant isolation, and can provision their own personal Slinky-enabled Slurm cluster within that namespace, complete with GPU access and workload scheduling logic.
Details: northwind.co and northwind.co

2. Mistral Compute (Managed Slurm) — Mistral AI's GPU cloud offers Slurm as a managed service with
queue management, priority tiers, and fair-share across teams, with topology-aware scheduling, multi-tenant by design, with policy enforcement across every cluster.
It also includes SSO/SCIM, RBAC mapped to Slurm accounts, and enterprise SLAs. See mistral.ai

3. Crusoe Managed Slurm — Note this is currently positioned as single-tenant rather than multi-tenant:
Crusoe Managed Slurm on CMK is intentionally focused: single-tenant GPU training clusters with a familiar Slurm interface, managed infrastructure, and the platform capabilities that come from running on CMK.
Worth watching if you need dedicated clusters rather than shared multi-tenancy: crusoe.ai

4. NVIDIA Slinky (open-source foundation, used to build your own service) — If you want to build a SLURMaaS yourself, Slinky is the underlying technology most vendors above use.
Slinky enables native Slurm cluster management on Kubernetes by representing all Slurm daemons as Kubernetes CRDs, and integrates with the NVIDIA GPU Operator and DRA/ComputeDomains for automated GPU management and topology-aware multinode scheduling.
developer.nvidia.com

5. vCluster (isolation layer for Slurm-on-K8s GPU platforms) — For providers wanting stronger tenant isolation than namespaces,
Slurm-on-Kubernetes is supported via Slinky integration, with vCluster virtualizing the K8s control plane to give every tenant a real API server without provisioning separate physical clusters.
vcluster.com

Recommendation: For a turnkey managed SLURMaaS product today, start with NorthWind (enterprise/cloud-provider focus) or Mistral Compute (if you want a full GPU cloud with Slurm built in). If building in-house, combine Slinky + vCluster on your own Kubernetes/GPU infrastructure.