We use server appliances to host websites and apps and provide technology services. Servers can be physical or cloud (virtual). Virtual servers are considered Cloud if they are hosted on a clustered IT infrastructure, known as "Cloud computing infrastructure". In such infrastructure, the appliances that provide processing power and data storage are separated into systems that are connected to the same computing network and feature redundancy of the computing resources.
Physical (Dedicated) servers and Cloud servers are two popular options for hosting applications, websites, and other services. With the development of computing devices, via the internet, embedded in everyday objects for the purpose of sending and receiving data, a new form of infrastructure service delivered from physical servers emerged - Edge Server Hosting.
What is Edge Server?
An Edge Server is a physical server that deals with instant data processing and is hosted as close as possible to "the Edge", where the data is collected. Edge servers could be an "Edge appliance" with minimal computing resources hosted in key points in major urban or peripheral areas that send that for processing to the closest possible data center. It could be also an "Edge Server" hosted in an Edge data center, that process the data received from the "Edge appliance".
So, when the term "Edge Server" is used we should read it as "physical dedicated server, hosted in an edge data center". When the term "Cloud Server" is used, we should understand it as a virtual computing instance that resides on a cloud infrastructure - AWS, Azure, Google Cloud, or any other cloud infrastructure provider.
Unlike the Cloud server, on the Edge one, there is an operating system installed directly on the physical server without virtualization technology. An Edge server could be also virtualized with any computing virtualization technology and made "Cloud-ready" and compatible with any major cloud computing platform such as VMware. Xen, Microsoft Hyper-V, Kernel-based VM, Linux Containers, Kubernetes, etc.
Like other physical servers, hosted in major metropolitan markets the Edge Servers are also a single-tenant infrastructure environment. Strictly, speaking any physical dedicated server can be an Edge one, if it has a 3ms - 5ms network latency to the targeted user market.
Like the Cloud-based servers, the Edge ones can be also scaled in RAM and storage to a certain degree, but they need to be shut down in order for a resource upgrade. In terms of CPU Edge servers are usually powerful hardware configurations with a lot of processing resources. However, their processing power cannot be scaled up, unless there is a clustered edge infrastructure in place, where multiple servers work as a system, that allows for new appliances to be plugged in.
Cloud Servers themselves could be "Edge Cloud Servers" if they are close to the edge market and if they reside on top of a clustered networking model of connected Edge Servers. Cloud servers in general are naturally scalable and their CPU, RAM, and data storage parameters can be scaled up or down on demand with or without being powered off, depending on the technology in use.
Edge servers are definitely a better choice for projects or applications that require the data to be processed as close to the originating source as possible, than the Cloud Servers. For this particular reason, in most IT scenarios, the major hyper-scalers line AWS or Azure, are not suitable for hosting edge IT infrastructure.