Are All Edge Devices “Servers?”

Over the weekend, a discussion broke out on Twitter between leading experts in the field of servers, IoT, gateways and single-board computers over the future of “edge computing.” This post is meant to summarize what was talked about for a broader audience because it is relevant to how we, as an industry, think and talk about “The Edge.”

The topic? What are the differences between an edge device and a server, should the industry consider these as being “The same thing” or are they different. Should devices which are meant to be deployed in fleets be treated differently than those meant to be deployed as servers?

Lets find out what key developers from the ecosystem said:

According to Red Hat Chief Arm Architect and current Nuvia VP of Software Jon Masters:

Edward Vielmetti, who manages special projects at Packet (a Bare Metal Server and On Premise Cloud Provider) said this:

David Tischler, founder of miniNodes weighed in:

You can view and participate in the full discussion here.

On a related topic, as has been reported in other publications, a community driven project is now well underway to certify the Raspberry Pi 4 as being “ServerReady.” Effectively what this would mean is that it is possible to buy $35 single-board computers which have similar software support to enterprise-grade server hardware.

Maintainers of the project recently posted a status update indicating that a comprehensive dashboard of progress is now available. The project is being organized via Discord currently with source code posted on GitHub.

Closing thoughts: What do you think? Is everything a server?

All opinions written here belong to the author and not those of their employer. Follow @rexstjohn on twitter because he posts on many hardware topics like robots, drones and edge computing.