A cloud server is made possible through virtualization. Management software called a hypervisor is installed on physical servers to connect and virtualize them: abstracting their combined resources and pooling them together to create virtual servers. These virtual resources can then be automated and delivered over the cloud for shared use in a single organization or across multiple organizations.
This approach is known as the infrastructure-as-a-service (IaaS) model. Organizations that employ IaaS don’t have to own and manage their own hardware; they can provision it from third parties that provide resources on demand via a public cloud. A common cloud server example is using a public cloud for temporary, seasonal or variable workloads that must be scaled up quickly as the need arises.
In some cases, however, cloud servers can also be configured as dedicated servers by a cloud provider. In this setup, sometimes called a bare-metal server, the provider dedicates physical cloud servers to one customer who may have specific performance or storage requirements.
To understand cloud server technology in more detail, read about cloud computing infrastructure and cloud management.