On this page, you can find an explanation of how to create GPU Virtual Machine, and instructions for other steps to manage Virtual Machines in Cloud Console.
GPU virtual machines (VMs) offer significant advantages for tasks that require intensive graphics processing or parallel computations.
š GPU services in the Cloud Console are currently accessible only for projects established in the Upper-Austria region.
Currently, our GPU services are exclusively available in the Upper-Austria region. We offer a range of GPU cards, including the H100 and L40S series, each available in multiple configurations to suit various performance needs:
ā¹ļø By default, these flavors are provided, but if you need a custom flavor, please contact our support team for assistance.
Choosing the appropriate GPU model allows you to enhance your applications for improved performance and efficiency within our cloud environment.
Creating a GPU VM involves a similar process to setting up regular Ubuntu or Windows VMs, but with a few additional steps specific to GPU Flavor configurations.
GPU VM Creation Process:
Name - specify a name for the virtual machine.
Source - choose the source for creating the VM: image, volume, or snapshot;
by default, “image” is pre-selected.
OS Platform - choose the operating system platform: Linux or Windows;
in this example we will create the Linux VM.
Linux Distribution - if Linux OS is selected, choose the desired distribution (e.g. Ubuntu, CentOS, etc.).
Image Version/Volume name/Snapshot name - based on the selected source, specify the corresponding image version, volume name, or snapshot name.
Choose Flavor Type - select GPU Flavor Type if you need a VM with GPU capabilities.
Choose Flavor - pick the type of GPU card and size for your new VM;
ā¹ļø available GPU flavors:
ā¹ļø If the list of GPU flavors is empty, your project is not currently associated with any GPU flavors.
š Please contact support to request access:
Key pair - required for Linux VMs; choose an existing SSH key (created on the SSH Keys page) or create a new one to connect to the VM via SSH.
Networks - select one or more networks to connect the VM to;
by default, “public” network is pre-selected.
Firewalls - choose what collection of network access rules will control the traffic to this VM;
by default, “default” Firewall is pre-selected;
š” Default Firewall allows access to the Internet from the VMs, but denies almost all access on the VMs from outside, except for objects belonging to the same default Firewall.
Tags - optional; use this field to assign tags to the VM.
Volume size (GB) - provide the preferred disk size for the VM;
valid range: 10ā1000 GiB;
minimal available size for Linux VMs - 10 GB; for Windows VMs - 50 GB;
by default, “50 GB” is pre-selected.
Delete Volume after VM deletion ā enable this option to automatically delete the volume when the VM is removed.Ā
After these steps, the newly created GPU VM will be added to theĀ Virtual MachineĀ pageĀ with the status ACTIVE.
The management actions for GPU VMs, such as Editing, Resizing, Associating/Disassociating a Floating IP, or Deleting, are identical to those for non-GPU VMs.
š” For additional details on these processes, refer to the Virtual Machines article.