What Is a Custom Server?
A custom server is a physical server configured to meet the specific requirements of a particular workload. Instead of choosing only from predefined server configurations, a business can select hardware and infrastructure parameters based on its computing, storage, networking, or application needs.
In simple terms, a custom server gives an organization more control over how a dedicated physical machine is configured.
Custom servers are useful when standard configurations do not provide the right combination of resources. Depending on the provider, customization may include CPU, RAM, storage, GPU, network configuration, bandwidth, IP addresses, operating system, and data center location. XLC offers custom dedicated servers with flexible hardware configurations that can be adapted to different workloads.
Custom Server Explained
A standard dedicated server typically comes with a predefined combination of hardware. A custom server allows that configuration to be adapted to the workload.
Custom does not necessarily mean that every component is manufactured specifically for one customer. In many cases, it means selecting available hardware and infrastructure options to create a configuration that better matches the intended workload.
A custom server may be configured around:
- CPU and processing requirements;
- RAM capacity;
- storage type and capacity;
- GPU requirements;
- network route and bandwidth;
- IP address requirements;
- operating system;
- data center location.
Key Takeaways
- A custom server is a physical server configured for specific workload requirements.
- It provides more configuration flexibility than a fixed server plan.
- Customization can cover hardware, networking, software, and deployment parameters.
- The available options depend on the provider and hardware inventory.
- Custom servers are useful when standard configurations do not match the workload.
How Does a Custom Server Work?
The process starts with identifying the workload requirements rather than choosing hardware first.
For example, a database may require substantial RAM and fast storage, while an AI workload may prioritize GPU capacity, memory, storage performance, and network throughput.
A typical custom server configuration involves:
- Define the workload — Identify applications, traffic levels, data volumes, and performance requirements.
- Select the hardware — Choose the available CPU, RAM, storage, GPU, and networking options.
- Choose the location — Select a data center based on latency, users, compliance, or connectivity requirements.
- Configure the software environment — Select the operating system and other required software parameters.
- Deploy the server — The provider provisions the selected configuration and makes the physical server available.
XLC’s server configuration workflow allows customers to select a region and adjust server, GPU, memory, storage, network, IP, and operating system parameters.
What Can Be Customized on a Server?
The available options vary between providers, but several components and infrastructure parameters are commonly considered.
CPU
CPU requirements depend on the applications running on the server. Workloads may prioritize core count, processing performance, or a particular processor architecture.
RAM
Memory capacity determines how much data and how many processes can be handled in memory at the same time. Databases, virtualization, analytics, and other memory-intensive applications may require substantial RAM.
Storage
Storage can be selected according to both capacity and performance requirements. NVMe SSDs, for example, can be appropriate for workloads that require high storage performance.
GPU
GPU-equipped configurations can support workloads that benefit from parallel processing, including AI and machine learning, computer vision, rendering, and other GPU-accelerated applications.
XLC currently offers dedicated AI server configurations with NVIDIA RTX 4090, RTX 5090, and H200 GPUs for AI training, inference, machine learning, and other data-intensive workloads.
Network and Bandwidth
Network requirements depend on traffic volume, user geography, application architecture, and data transfer needs.
A custom configuration may include selecting a network route and bandwidth level appropriate for the workload.
Operating System
The operating system should match the application’s software stack and operational requirements. Depending on the provider, customers may be able to choose between supported Linux distributions and Windows Server configurations.
Custom Server vs Standard Dedicated Server
A custom server and a standard dedicated server are both dedicated physical infrastructure options. The main difference is how the server is configured.
A dedicated server describes the tenancy model: the physical machine is allocated to one customer.
A custom server describes the configuration: its hardware and infrastructure parameters are selected or adjusted for a specific workload.
This means a dedicated server can be either a predefined configuration or a custom configuration.
| Characteristic | Custom Server | Standard Dedicated Server |
|---|---|---|
| Hardware configuration | Tailored to workload requirements | Predefined |
| Component selection | Greater flexibility | Limited to available configuration |
| Deployment | May require additional configuration | Usually faster |
| Specialized hardware | Easier to accommodate | Depends on available plans |
| Best suited for | Specific or complex workloads | Common workloads with predictable requirements |
Businesses looking for ready-to-deploy infrastructure can use dedicated servers, while custom configurations are more suitable when predefined hardware does not match the workload.
Custom Server vs Cloud Server
A custom physical server and a cloud server provide different approaches to infrastructure.
A custom server provides dedicated physical hardware with direct control over its configuration. A cloud server typically provides virtualized computing resources that can be provisioned through a cloud platform.
| Characteristic | Custom Server | Cloud Server |
|---|---|---|
| Infrastructure | Physical hardware | Typically virtualized resources |
| Hardware control | High | Limited at the physical hardware level |
| Resource isolation | Dedicated physical resources | Depends on cloud architecture |
| Scaling | Requires hardware changes or additional servers | Usually more flexible |
| Custom hardware | Available depending on provider | Usually limited to provider-defined options |
| Best suited for | Specialized or predictable workloads | Flexible and changing workloads |
The appropriate model depends on the workload, scaling requirements, budget, performance needs, and level of hardware control required.
When Do You Need a Custom Server?
A custom server can be useful when a standard configuration does not provide the right combination of resources.
Specialized Hardware Requirements
Some workloads require a specific combination of CPU, RAM, storage, GPU, or networking resources that is not available in standard plans.
AI and Machine Learning
AI workloads may require dedicated GPUs, large amounts of memory, fast storage, and high-throughput networking. XLC’s AI servers are configured for AI training, inference, machine learning, and other GPU-intensive workloads.
Large Databases
Database workloads can require substantial RAM, fast storage, and consistent processing performance.
High-Traffic Applications
Gaming platforms, eCommerce services, streaming platforms, and other high-traffic applications may require dedicated resources and network configurations matched to their traffic profiles.
Specialized Infrastructure
Organizations may need specific hardware or deployment locations because of application requirements, data residency, latency, or operational constraints.
Benefits of Custom Servers
Workload-Specific Configuration
Resources can be selected according to the actual requirements of the application instead of forcing the workload into a predefined configuration.
Dedicated Physical Resources
A custom dedicated server provides dedicated CPU, RAM, storage, and other physical resources for one customer. XLC describes its dedicated infrastructure as single-tenant bare metal.
Hardware Control
Organizations have greater control over the physical configuration and software environment than they typically have with virtualized infrastructure.
Support for Specialized Workloads
Custom configurations can accommodate workloads that require specific combinations of computing, storage, GPU, and networking resources.
How to Choose a Custom Server Configuration
Start with the workload requirements rather than the available hardware.
Consider:
- CPU requirements — Determine the processing capacity required by the application.
- Memory requirements — Estimate how much RAM is needed for applications, databases, or other workloads.
- Storage requirements — Consider both capacity and storage performance.
- GPU requirements — Determine whether the workload benefits from GPU acceleration.
- Network requirements — Estimate bandwidth, traffic volume, and user geography.
- Location — Choose a data center region based on latency, users, compliance, and connectivity.
- Growth — Account for expected increases in traffic, data, and compute requirements.
The goal is not to select the largest configuration available, but to build a server that provides the resources the workload actually needs.