Can You Define the Edge?
In a new Thoughts paper, Infrastructure Mason Smarak Bhuyan argues that we don’t have, but we desperately need, a widely accepted definition of Edge.
For many years, confusion about what the cloud actually is was pervasive, even among those in the IT industry. Now, of course, there is a widely accepted definition of the cloud. But with Edge, we are at the same point today as we were with cloud a decade ago – many different ideas, and no real consensus, about what exactly Edge is.
Why definition matters
Why not just channel Justice Potter Stewart and say “We know it when we see it”? A widely accepted definition will be essential as we work together to confront the challenges that will arise as more work is done at the Edge.
As one Infrastructure Masons Advisory Council member said, “There will soon be two times more compute done outside of what we call data centers – at the edge, distributed compute, devices. And that has to be done in places where there might not even be roads.” Most worrisome is the people aspect. “It is going to happen but who has the skill sets to invent, deliver, and keep it operational?”
“A definition will be essential as we work together to confront the challenges that will arise as more work is done at the Edge.” – Click to tweet
A working definition
Troubled by the lack of a widely accepted definition, Infrastructure Mason Smarak Bhuyan took it upon himself to write a Thoughts paper, Edgy About the Edge, to put forth a working definition. This is how Smarak defines Edge:
“An Edge location is a computing enclosure/ space/ facility geographically dispersed to be physically closer to the point of origin of data or a user base. In other words, for an Edge to exist there must be a hub or a core; therefore, dispersion of computing to the periphery would qualify as ‘Edge computing’ and the physical enclosure/ space/ facility can be defined as the ‘Edge facility’.”
Based on Smarak’s definition above, which of the following should qualify as the Edge:
- A large colocation facility in a tier two city that extends the reach of the internet outside hubs like Dallas, Chicago, or a smaller facility that a Content Distribution Network (CDN) owns across the globe.
- Cylindrical lights out rack chambers (150 kW) installed under cellular towers or micro datacenters (300 kW), which can be installed anywhere.
- An autonomous car with massive computing platforms or a datacenter on a cruise ship floating on the edge of the Antarctic.
(You’ll have to go to the Thoughts paper to check your answers; refer to page 11.)
“For an Edge to exist there must be a hub or a core.” – Click to tweet
4 use cases for edge computing
- Content – “For example, a content delivery network will download a popular movie from Netflix’s central archives in Los Angeles to a Dallas Edge data center one time. That long-haul download of the 6 GB movie is expensive, but distributing the movie to all the Dallas-Fort Worth viewers from the Dallas Edge data center is much cheaper.” (Get more details on page 8 of the Thoughts paper.)
- Localization – “Larger organizations may be willing to invest in local processing of data for latency reasons. These are smaller deployments with computing storage and network capabilities.” (Get more details on page 9 of the Thoughts paper.)
- IoT – “IoT devices are sensors with a gateway which aggregate data and connect to the internet or WAN for transit…Depending on the type of workload, criticality needs and application latency, the computing needed can be local with a local database for processing.” (Get more details on page 9 of the Thoughts paper.)
- Next-generation workloads – “Next-generation application workloads such as augmented reality and virtual reality, drone footage, and autonomous cars will require massive near-real time computation…If these workloads can be processed close by at Edge facilities, the user experience will be dramatically better.” (Get more details on page 10 of the Thoughts paper.)
“An Edge location is geographically dispersed to be physically closer to the point of origin of data or a user base.” – Click to tweet
As Edge deployments become more common, the organizations running them will have to confront a range of challenges, including managing widely distributed assets and security at lights out facilities. Perhaps most challenging: efficiency, because the number of Edge facilities will likely dwarf the number of core data centers, and may not benefit from the same efficiencies that come with massive scale.
What do you think?
IM Thoughts papers are designed to be conversation starters. So, what do you think? Do you agree with Smarak’s definition of Edge? Or would you change it? Share your thoughts in the comments below or on LinkedIn (members only) and Twitter.