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

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

1. NorthWind-Powered SLURM-as-a-Service (SLURMaaS)
NorthWind's SLURMaaS
enables organizations to deliver fully managed, multi-tenant SLURM environments for high-performance computing (HPC) workloads
. It
modernizes SLURM by integrating it with Kubernetes through Project Slinky, allowing providers to expose SLURM job scheduling as a cloud-like, on-demand service
. Key features include
traditional SLURM scheduling with Kubernetes orchestration, dynamic GPU allocation, and multi-tenant security, launching clusters instantly, optimizing utilization across SLURM jobs automatically, and secure workload separation with per-tenant tracking
. Architecturally,
each user (e.g., a research scientist, ML engineer, or analytics team) operates in their own NorthWind-managed Kubernetes namespace, ensuring strong tenant isolation
, and users can
provision their own personal Slinky-enabled Slurm cluster within their namespace, complete with GPU access and workload scheduling logic
.
→ northwind.co
→ northwind.co

2. vCluster Platform (Certified Stacks)
vCluster targets AI/GPU clouds by running SLURM inside isolated virtual Kubernetes clusters. Its
Certified Stacks deliver Ray, Run:AI, and Slurm on your GPU racks — no months of integration
, and it's positioned as
a powerful solution for any multi-tenant Kubernetes use case requiring strong, cost-effective isolation... particularly well-suited for the demands of AI/GPU clouds
.
→ vcluster.com

3. NVIDIA Slinky (SchedMD, now part of NVIDIA) + GPU Operator
The underlying technology powering the above platforms:
Slinky, developed by SchedMD (now part of NVIDIA), enables native Slurm cluster management on Kubernetes by representing all Slurm daemons as Kubernetes Custom Resource Definitions, supporting full Slurm lifecycle orchestration and high availability
. It integrates with
the NVIDIA GPU Operator and DRA/ComputeDomains allowing automated GPU management, topology-aware multinode scheduling, and per-job GPU monitoring, supporting advanced NVIDIA architectures like GB200 NVL72
.
→ developer.nvidia.com

For a bare-metal, non-Kubernetes approach, plain SLURM with NVIDIA MIG partitioning also supports multi-tenancy:
Multi-Tenant Support is ideal for cloud service providers and data centers that host services for multiple customers, each requiring different levels of GPU resources
.
→ techcommunity.microsoft.com

NorthWind and vCluster are the closest to true "as-a-Service" commercial offerings, while Slinky/NVIDIA GPU Operator is the open infrastructure layer they build on.