GPU Dedicated Server Overview
A GPU Dedicated Server gives you access to a physical server with dedicated GPU resources assigned to your service.
This section helps you understand how GPU Dedicated Server ordering, management, billing, and troubleshooting work in PerLod.
What a GPU Dedicated Server is
Unlike a standard Dedicated Server, a GPU Dedicated Server includes one or more dedicated GPUs as part of the hardware setup.
This makes it suitable for workloads that need more than CPU-only processing, especially when performance depends on GPU acceleration, parallel computation, or GPU-enabled software.
Common use cases
GPU Dedicated Servers are commonly used for:
- AI and machine learning workloads
- model training
- inference workloads
- rendering tasks
- GPU-based compute jobs
- research and testing environments
- high-performance application workloads
The right server depends on your GPU requirements, system resources, software stack, preferred location, and deployment goals.
How the GPU Dedicated Server flow works
The general flow is:
- start from the main GPU Dedicated Server page or a location-based page
- filter and compare available GPU server options
- select a server
- configure the order
- review the order
- complete payment
- manage the active service in Dash
The website is used for browsing and ordering. Dash is used for managing the server after provisioning.
What is covered in this section
This section includes documentation for:
- ordering a GPU Dedicated Server
- managing the service in Dash
- common service control actions
- billing and lifecycle handling
- troubleshooting common issues
Before you order
Before choosing a server, it helps to review:
- GPU model requirements
- number of GPUs if needed
- CPU requirements
- memory requirements
- storage capacity
- storage type
- bandwidth expectations
- preferred location
- operating system requirements
- workload type
Choosing the right server early helps reduce unnecessary changes later.