Features and Benefits

Features

• Power operations: Manage your Cloud GPU VMs with standard power operations, including stop, start, and restart capabilities.

• STOP deallocates the Compute unit of your instance and returns the GPUs, CPUs, and RAM to the hardware pool. Block Storage remains provisioned, allowing you to resume work after restarting the instance. Billing will be paused when you stop the Cloud GPU instance.

Note: Deallocated resources may be used for other IONOS Cloud users' workloads. Depending on availability, an instance may fail to start if all resources are currently in use.

• START provisions the Compute unit and allocates the GPUs, CPUs, and RAM as specified in your VM template configuration. The Block Storage volume is remounted to the VM instance, restoring access to your data and configurations. Billing for the instance will resume immediately.

• REBOOT executes a power cycle of your Cloud GPU instance. This operation does not deallocate resources, as the instance remains active. Billing is not paused during this operation.

• Server configuration: Provision servers using pre-configured templates with Cloud GPU VM specifications, or customize storage by attaching additional volumes that scale independently.

• Operating system support: Cloud GPU VMs currently support Linux-only virtual machines.

• Auto availability zone: Cloud GPU VMs deploy automatically across availability zones for optimal resource distribution.

• Volume operations: The Cloud API and the DCD supports upscaling additional volumes and attaching or detaching volumes. The first connected volume (storage volume) cannot be detached or upscaled.

• Availability: It is currently available only in the Frankfurt de/fra/2 data center.

Benefits

• Complete control and flexibility: Create and delete your Cloud GPU VMs on demand, giving you full control over resource usage and costs while maintaining uninterrupted workflow management.

• Reliable Linux-based infrastructure: Leverage the stability and performance of Linux virtual machines, optimized for Cloud GPU VM workloads and available through a streamlined Cloud API and through the DCD for efficient automation and integration.

• Scalable storage architecture: Expand your storage capacity on the fly by attaching additional volumes or increasing their size, allowing your infrastructure to grow seamlessly alongside your projects without downtime.