Skip to main content

Getting Started

Deploy and connect to your first Hyperstack GPU virtual machine. Follow the steps below to configure the VM, deploy it, and connect over SSH.

Looking for Something Else?

This guide deploys a GPU virtual machine. To deploy a Kubernetes cluster instead, see Getting Started with Kubernetes. To run or fine-tune models without provisioning infrastructure, see Get Started with AI Studio.

Before you begin​

You'll need:

Deploy and connect to a VM​

  1. Open the Deploy New Virtual Machine page​

    On the Virtual Machines page, click Deploy New Virtual Machine.

  2. Choose a flavor​

    The flavor determines your VM's GPU, vCPU, RAM, and storage. 1x L40 is recommended for a first deployment, or choose the flavor that best fits your workload. See all available flavors.

  3. Create an environment​

    In the Environment section, click Create New Environment, give it a name, and save. An environment groups VMs, SSH keys, and other resources together. The recommended region for your selected flavor is chosen by default. See Environments for details.

  4. Select an OS image​

    In the Select OS Image section, the default Ubuntu image Server 22.04 LTS R535 CUDA 12.2 is pre-selected. It comes CUDA-ready, so no additional driver or toolkit setup is needed. To pick a different OS or CUDA version, see Images.

  5. Create an SSH key​

    Click Create New SSH Key, then choose how to add your key:

    To generate a new key, select the Generate SSH key tab, choose the environment you just created, enter a name, and click Generate. Hyperstack creates the key pair and downloads the private key file to your computer. Save the path to your private key, you'll need it to connect to your VM over SSH.

  6. Enable SSH access​

    In the SSH Access section, toggle the option to Enabled so you can connect to your VM over SSH. This automatically adds a firewall rule allowing incoming traffic on TCP port 22.

  7. Assign a public IP​

    In the Public IP Address section, toggle on Assign Public IP. This gives your VM an internet-reachable IP address so you can connect to it over SSH from your local machine. Learn more about Public IPs and Internet Accessibility.

  8. Deploy​

    Click Deploy to begin provisioning. The VM is ready to use once it enters the ACTIVE state.

    Boot time

    Smaller flavors like 1x L40 usually reach ACTIVE in a few minutes. Larger configurations like 8x H100-80G-PCIe-NVLink can take longer. Billing does not start until the VM is ACTIVE (see VM states and billing for details).

  9. Connect to your VM over SSH​

    Once your VM is ACTIVE, open your local terminal and run the command below for the OS image you selected. Replace:

    Connect over SSH
    ssh -i [path_to_ssh_key] ubuntu@[vm_ip_address]

    For example:

    ssh -i ~/Downloads/my-key.txt [email protected]
    Troubleshooting SSH key permission errors

    If SSH rejects the key with Permissions ... are too open or Permission denied (publickey), run the command below to restrict the key file (replacing [path_to_ssh_key] with your actual key path):

    chmod 600 [path_to_ssh_key]

    Then re-run the SSH command.

    Once the connection succeeds, you'll see the Ubuntu welcome banner and a shell prompt:

    Example Ubuntu welcome banner
    Welcome to Ubuntu 22.04.5 LTS (GNU/Linux 6.8.0-40-generic x86_64)

     * Documentation:  https://help.ubuntu.com
     * Management:     https://landscape.canonical.com
     * Support:        https://ubuntu.com/pro

     System information as of Thu Jan 15 12:00:00 UTC 2026

      System load:  0.05               Processes:             350
      Usage of /:   15.0% of 100.00GB  Users logged in:       0
      Memory usage: 0%                 IPv4 address for ens3: 10.0.0.42
      Swap usage:   0%

    ubuntu@my-vm:~$
Your VM is ready

Start running your workload. Common uses include AI training and inference, 3D rendering, scientific and engineering simulation, and data analytics.

Managing your VM

When you're finished, hibernate to reduce charges or delete the VM if you no longer need it. See VM Status and State Management for lifecycle options.

Next steps​

Learn more about Hyperstack and get more out of the platform:

Virtual Machines

Explore advanced configuration and lifecycle options for your VM like images, snapshots, hibernation, and more.

Storage

Learn about the storage options available for your VM, including ephemeral, persistent, and shared volumes.

Network

Configure your VM's networking, including SSH, firewalls, public IPs, and high-speed connectivity.

Kubernetes

Scale beyond a single VM with on-demand GPU clusters for orchestrated, multi-node workloads.

On this page