Spin up the compute power you need the moment you need it. With CPU- and GPU-based instance (virtual server) options, portvMind Compute scales just as easily from a single-server test environment to a multi-node production cluster. When your resource demand changes, your infrastructure doesn’t keep you waiting.
Compute is a virtual server service that gives you access to processing power (CPU, RAM and, when needed, GPU) without buying a physical server. On portvMind, every instance draws from isolated resource pools on an OpenStack-based virtualization layer — heavy use by a neighboring instance does not affect your performance.
In practice it works like this: you log in to the console, choose the operating system and instance size, and within a few minutes you have a server you can connect to via SSH or RDP. When your needs change you scale resources up, and when you don’t need them you shut them down. Because metering is per minute, if you run a test environment for forty minutes and shut it down, you pay only for those forty minutes.
Your servers are located in the Istanbul or Ankara region. Which region you choose is your decision, and your data does not leave that region.
Your server is ready with a few choices in the console. No hardware procurement, installation or configuration process.
From general-purpose workloads to model training, you choose the instance family that fits your scenario. Cards dedicated to AI workloads are in the GPU / AI Compute service.
The billing unit is the minute; there is no hourly or daily rounding. Test environments you shut down after hours turn directly into savings.
Combined with Block Storage, you get a performance profile where disk I/O is not a bottleneck.
Runs in the Istanbul and Ankara facilities; latency and data sovereignty are solved at the same time.
Take a snapshot of an instance and derive new servers from it, standardizing your environments.
| Feature | Detail |
|---|---|
| Instance families | General purpose · memory optimized · GPU-backed |
| Operating systems | Ubuntu, Rocky Linux, Debian, Windows Server · custom image upload |
| Storage | Full-flash NVMe system disk; expandable with Block Storage |
| Network | VXLAN-based private network, security groups, Floating IP |
| Regions | TR-IST-01 (Istanbul) · TR-ANK-01 (Ankara) |
| Metering | Per minute |
The full list of vCPU, memory and disk combinations can be seen in the calculator.
For detailed technical specifications and release notes, see Documentation
Enter your configuration and see the total cost instantly.
Metering: Per minute
For predictable workloads you can lower your unit cost with reserved commitment options.
Your resources are metered for as long as they are running — you never pay for time you don’t use. In the console you can define a monthly or period-specific budget and get alerts when your spending reaches the 50%, 80% and 100% thresholds. Turn on overrun forecasting to see in advance whether your current consumption rate will take you over budget by the end of the period.
Alerts arrive by e-mail; with webhook support they can also be sent to Slack, Microsoft Teams or your own application.
Budget alerts send notifications; they do not stop resources automatically. You decide which resource to scale down.
The full pricing modelAfter you complete the configuration in the console, your instance is usually up and running within a few minutes.
As long as your server is running, it is metered per minute; there is no hourly or daily rounding. For long-term, predictable workloads you can lower your unit cost with reserved commitment options.
Metering stops for Compute resources. The disk space attached to the server continues to be metered because your data is preserved.
Yes. When your resource needs change you can move to a larger type; the downtime varies with the operating system and configuration.
Yes. You can upload images in standard formats and create servers from them.