Adding a VM to a GPU cluster
After creating a GPU cluster, you can add new VMs with GPUs to it. It is not possible to add already existing VMs to a GPU cluster.
VMs in a GPU cluster must have 8 GPUs each and can use either of these platforms:
- NVIDIA® H100 NVLink with Intel Sapphire Rapids (Type A) (
gpu-h100
) - NVIDIA® H100 NVLink with Intel Sapphire Rapids (Type B) (
gpu-h100-b
) - NVIDIA® H100 NVLink with Intel Sapphire Rapids (Type C) (
gpu-h100-c
)
The NVIDIA H100 platforms provide identical performance and capabilities.
We recommend to create a cluster with the VMs in the same network.
For the cluster VMs to interact properly, we recommend using a security group that allows unrestricted traffic within itself. For example, the default security group meets this requirement.
For configurations that can be used in GPU clusters, see Configurations of VMs with GPUs.
While creating a VM, under Computing resources, in a GPU cluster, select an existing GPU cluster or click Create button to create a new one.
export YC_GPU_CLUSTER=$(ncp compute gpu-cluster list --format=json | jq -r .[].id)
ncp compute instance create --name node-gpu-test \
--create-boot-disk size=64G,image-folder-id=standard-images,image-family=ubuntu-2204-lts-gpu-cluster-cuda12,type=network-ssd \
--ssh-key=$HOME/.ssh/id_ed25519.pub \
--gpus 8 --cores 160 --memory=1280G \
--network-interface subnet-name=default-eu-north1-c,nat-ip-version=ipv4 \
--platform gpu-h100 \
--gpu-cluster-id=$YC_GPU_CLUSTER
This command uses a Nebius AI Marketplace image adapted for GPU clusters, Ubuntu 22.04 LTS for NVIDIA® GPU clusters (CUDA® 12) (--create-boot-disk image-folder-id=standard-images,image-family=ubuntu-2204-lts-gpu-cluster-cuda12
). You can find other images for GPU clusters on Marketplace.