> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nebius.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Slurm and Soperator in Nebius AI Cloud

Soperator is an [open-source solution from Nebius](https://github.com/nebius/soperator) that allows you to consolidate Slurm and Kubernetes® into a single infrastructure. You can manage nodes by using standard Kubernetes resources and run machine learning experiments by using Slurm.

You can use Soperator in the following different environments:

* [Managed Service for Soperator](/slurm-soperator/deploy/overview#managed-service-for-soperator): Nebius AI Cloud service
* [Pro Solution for Soperator](/slurm-soperator/deploy/overview#pro-solution-for-soperator): deployed by Nebius solution architects on Nebius AI Cloud
* [Self-deployment on a Managed Service for Kubernetes cluster](/slurm-soperator/deploy/overview#self-deployment-in-nebius-ai-cloud)
* [Other cloud platforms and on-premises](/slurm-soperator/deploy/overview#self-deployment-on-other-platforms-and-on-premises)

*InfiniBand and InfiniBand Trade Association are registered trademarks of the InfiniBand Trade Association.*

<CardGroup cols={2}>
  <Card title="About Slurm and Soperator" icon="link" href="/slurm-soperator/overview/why-slurm-soperator">
    Get acquainted with key Soperator features
  </Card>

  <Card title={`Deploying Soperator clusters`} icon="link" href="/slurm-soperator/deploy/overview">
    Choose a deployment method that fits your use case best
  </Card>

  <Card title={`Creating clusters in Managed Service for Soperator`} icon="link" href="/slurm-soperator/managed-soperator/manage">
    Get started with Slurm and Soperator in Nebius AI Cloud with minimum effort
  </Card>

  <Card title={`Connecting to Slurm nodes`} icon="link" href="/slurm-soperator/clusters/connect">
    Connect to login and worker nodes, so you can start managing machine learning workloads
  </Card>

  <Card title="Running batch jobs" icon="link" href="/slurm-soperator/jobs/index">
    Define, configure and launch your workloads in Slurm
  </Card>

  <Card title="Managing jobs and queue" icon="link" href="/slurm-soperator/jobs/manage">
    View and control current and historical jobs
  </Card>

  <Card title="Running the all-reduce NCCL test" icon="link" href="/slurm-soperator/jobs/examples/nccl-all-reduce">
    Check the NVLink and InfiniBand™ performance between GPUs on one or multiple nodes
  </Card>

  <Card title="Managing users" icon="link" href="/slurm-soperator/users/manage">
    Create users, so they can connect to Slurm nodes
  </Card>

  <Card title="Monitoring job and node statuses" icon="link" href="/slurm-soperator/monitoring/statuses">
    Get up-to-date information about Slurm jobs and nodes
  </Card>

  <Card title="Ephemeral nodes" icon="link" href="/slurm-soperator/capacity/ephemeral-nodes">
    Reuse reserved GPU capacity across training and inference workloads
  </Card>
</CardGroup>