- Edge computing is a real thing, but distorted and extended beyond reasonable use cases by FOMO.
- Smart edge computing plays are not generalized, but specialized, and they do not play on hype.
The first conversations around the concept of edge computing were both interesting and enlightening. The basic idea was that compute resources needed to be closer to the actual workload in situations where real-time or very near-real time decisions need to be made. Latency could not be tolerated, so cloud or even corporate data centers were out of the question. Examples given were automated materials handling, manufacturing, and – of course, technology marketers’ favorite old trope – self-driving/automated vehicles. All but the last example sounded perfectly reasonable and lined up with customer needs, both today and tomorrow.
Then the story changed. ‘Fear of missing out’ (FOMO) entered the chat, and a perfectly reasonable technology term was suddenly the ‘next hot thing’ and everybody had to play, regardless of whether or not it made sense. Market forecasts have been generated, big white papers written, and coverages initiated, all singing the praises and the billions of dollars to come. The lyrics of this new song are unfamiliar, but the underlying tune has been played over and over again, countless times. Remember blockchain? Technology trends, especially those that are solutions looking for problems, are an issue. FOMO and marketing make everyone talk about technologies that are good, but will not change the world the way proponents claim.
Now edge computing is getting blown up and distorted beyond all recognition. Smartphone in your pocket? Edge computing. Data center somewhere not in a hyperscale data center? Edge computing. Got a rack of 10-year-old equipment designed to run a specific system in a closet somewhere? Edge computing. Got a computer in your car? You guessed it… edge computing. Cloud computing companies are selling ‘edge cloud’ solutions. But classifying every IT closet in the world that isn’t in HQ as an ‘edge data center’ is – well, frankly – marketing garbage. Did we all collectively forget that this is how data processing was done before cloud computing, which on the grand scale of things isn’t even that old?
The original intent of edge computing was to solve a latency issue or presort data before uploading to a cloud, used locally, or to the data center for storage or further processing. Everybody is sure, fanatically sure, that edge computing is going to be big, huge, a game-changer for customers, service providers, and vendors alike. But when you press for use cases, there is a lot of hand-waving that ends with somebody mentioning automated cars. Real use cases and proof that a given vendor or SP can deliver them are what’s missing from the current edge computing craze.
GlobalData is getting more and more inquiries from our customers about edge computing, most of them frantically trying to relabel or tweak current offerings into edge computing because of FOMO. But the best advice is: instead of trying to fulfill some broad, nebulous concept that is daily being marketed a hundred different ways, concentrate on concrete use cases. Specialize in doing a few edge computing things well. Don’t let FOMO or your sales organization push your company into a broad, top-level marketing category. Make it a dropdown on your main website that leads customers directly into your tuned and specialized solutions in specific vertical markets. Don’t indulge in big but empty marketing around how awesome edge computing is. Get right to what you can do and how. Success there will lead to more customers both in those verticals and outside coming to your organization to solve actual edge computing use cases. Success is in proving you can solve problems, not catering to the ‘next big thing’ without concrete results.