In 2014 when I first really started to dive into the opportunity of Edge, I believed then, that Edge would be a large market, how large, I didn’t attempt to predict. In January of 2019 I wrote The Edge Marketplace - Bigger than the Internet. Now what I’m seeing is making me believe my own prognostications even more fiercely than before.
Of Course Mark believes it, he founded an Edge oriented business
It’s true, before you read any further, you must first consider the source. In this case, this author believes so much that he founded a company to facilitate edge development and growth, so yes, he (I) does have an ax to grind or as they also say, a horse in the race.
Four (4) of the things I’m seeing that are still true
There isn’t a single “Killer app”
Most folks have been or are still, looking for the Killer App for the Edge to “really get going”. I’ve said from the early days of edge that there are killer verticals, but there is unlikely to be a killer app. A killer vertical is something like manufacturing or healthcare vs a killer app which I would characterize as something like a Fitbit or shop floor automation. A killer app is also characterized by the assumption that it will, by itself, drive or fund technology deployment at the edge. On the other hand, Killer verticals are real and are many, from blockchain to bulldozers and from data sovereignty to data analytics with AI. Several leading industry firms including Gartner have indicated that this is the year when investment in edge by the enterprise will really take off and I heartily agree. In fact Gartner says enterprises will increase their investments in Edge this year by 76% on average from $262K to $462K
Just because you don’t recognize it doesn’t mean “it” (edge) hasn’t arrived
Those of us in the data center or infrastructure space often look for edge to show up like the Kardashian’s entourage, something you couldn’t miss. In the consumer space, folks expect edge to announce itself when they use technology “thank you for using this facility, which was enabled via an edge solution”. I’m saying the edge has already arrived in important and critical ways, we just don’t recognize it as such. One of the problems for the infrastructure geeks like me relates to point 1 above, without some huge killer app that by itself requires hundreds of new data centers distributed globally, edge seems to not be happening.
Edge is more likely to be dozens to hundreds of locations with 1-10 servers
If 1000 ants attacked you individually over the course of several days, you probably wouldn’t think “Oh my god, they’re taking over the world”. However, if 1000 ants all attacked you at the same time you would likely take notice. Edge deployments are more like the ants attacking one to three at a time. Instances are being deployed for thousands of different use cases in individual locations and in thousands of locations, but they arrive by individual UPS delivery, not via a train or even a semi.
There isn’t one flavor of edge infrastructure and there likely won’t be
The battle for the edge is far from over, in fact, the options for how and where to deploy technology solutions for edge are still being explored. What I can say that is true is that time to value, cost effectiveness and agility will continue to rule for decision makers.
One thing I’m seeing that’s different
Edge doesn’t always mean nanoVMs, containers or Functions-as-a-Service (FaaS or Serverless)
Deploying at the edge will often mean the use of technologies that allow for small footprint and efficient use applications, however, the opportunities are too numerous to wait for every application to be rebuilt as microservices. We will for some time (similar to mainframes and on-premises IT) see things like VMs etc at the edge. When you have a defined opportunity for an IT solution, even a solution that’s less than perfect is often better than not having one at all.
Edge marches on
The opportunities to leverage edge for new services, saving lives and differentiating with customers continue to expand in number, but the market can still use help in how it develops and expands. Also, arguments about ownership (hyperscalers vs others vs a mix) will continue. I personally hope we have a new set of players to add options to what the hyperscalers bring to the table. As is often the case these days, the developer experience will be a big part of who wins. In the meantime, if you have a motivation (use case) for being at the edge, don’t hesitate to pursue it. Yes, it’s noisy space and the solution landscape is frightening in its complexity and diversity, but that shouldn’t stop you. “S/he who dares, wins.” My next Edge blog will likely be 'Edge Evolution - The Next 7 Years'