Find self-service SLURM-as-a-Service options for AI workloads.

Here are several self-service SLURM-as-a-Service options tailored for AI workloads, ranging from managed cloud offerings to hybrid and on-premises solutions:

• Google Cloud Cluster Director

A managed service that automates provisioning and control of Slurm and Kubernetes clusters. Users can deploy AI or HPC clusters using the control plane, CLI, or API within Google Cloud, benefiting from reduced latency and scalable infrastructure. (cloud.google.com)

• TensorWave Managed Slurm

Provides fully managed Slurm clusters optimized for GPU-heavy AI training and inference workloads. Features include topology-aware workload mapping and integration with Kubernetes for unified orchestration. (tensorwave.com)

• NorthWind-powered SLURM-as-a-Service

A cloud-like, multi-tenant solution offering on-demand Slurm cluster provisioning via portal or API, with secure isolation and built-in governance—ideal for organizations needing flexible, self-service access. (northwind.co)

• Nscale Platform Services

Offers managed, enterprise-ready Slurm scheduling orchestrated atop Kubernetes. This supports multi-tenancy and rich observability, allowing AI teams to shift from HPC to AI workloads smoothly. (nscale.com)

• Crusoe Managed Slurm

Enables provisioning of complete Slurm clusters on Crusoe Cloud via a simple UI or CLI. It delivers GPU-optimized infrastructure, topology-aware scheduling, shared storage, and a unified Slurm experience. (docs.crusoecloud.com)

• Lambda Managed Slurm (1‑Click Clusters)

Offers both unmanaged and fully managed Slurm deployments on GPU clusters (e.g. HGX B200/H100). The managed option includes support, daemon monitoring, job history, patching, and more. (lambda.ai)

• RedFort Tech Managed SLURM

A SLURM-as-a-Service solution with preconfigured controllers and GPU partitions (H100, B200, A6000), fair-share scheduling, elastic scalability, and automation for rapid deployment in the cloud. (redforttech.us)

• Waldur (open-source)

A self-service cloud and HPC orchestration platform that automates Slurm and OpenStack deployments. It includes billing, user self-service, and integrated support-desk features. (waldur.com)

Each of these services supports self-service provisioning and streamlines Slurm-based AI workflows, so you can choose based on your preferred environment—cloud-native, hybrid, or on-premise.