NVIDIA® Network Operator

The NVIDIA^® Network Operator is a Kubernetes application designed for managing and optimizing software components for networking between NVIDIA GPUs in the cloud. The operator automates many tasks related to network setup, including the configuration of high-performance networking features like RDMA (Remote Direct Memory Access) and GPUDirect, which are crucial for applications requiring low latency and high throughput. This tool is particularly beneficial for environments where NVIDIA GPUs are deployed for compute-intensive tasks, as it ensures that the network can support the high data transfer demands of such applications.

You can deploy the NVIDIA Network Operator in your Nebius AI Managed Service for Kubernetes clusters using this Marketplace product. Your cluster must have a node group attached to a Compute Cloud GPU cluster.

Before installing this product:

1. Create a GPU cluster in Compute Cloud.
2. Create a Kubernetes cluster and a node group in it. When creating the group, select the created GPU cluster for it.
3. Install kubectl and configure it to work with the created cluster.

To install the product:

1. On the cluster page in the management console, go to the Marketplace tab, select the product, and click Install.

2. Configure the application:

   • Namespace: Select a namespace or create a new one.
   • Application name: Enter an application name.

3. Click Install.

4. Wait for the application to change its status to Deployed.

5. To check that the NVIDIA Network Operator is working, check that its pods are running:

   kubectl get pods -n <namespace>
   

• Automating management of software components for GPU networking in Kubernetes clusters.
• Building fast infrastructures for high-performance computing (HPC) and AI workloads.

Nebius AI does not provide technical support for the product. If you have any issues, please refer to the developer’s information resources.