Who Wins Edge Computing?

Mobile operators hope they can play a significant role in edge computing. Of course, to the extent that  edge computing revenue is expected to become significant, lots of other firms in the computing ecosystem will try to grab a position as well. Of course, edge computing will continue to be dwarfed, over the next five to 10 years, by centralized cloud computing.

Still, many Internet of Things and immersive and interactive use cases “will flip the center of gravity of data production and computing away from central data centers and out to the edge,” say some researchers at Gartner.  

source: 451 Research

“Instead of continued growth of mega data centers, compute and storage will move toward the edge, due to the Internet of Things and new user/machine interfaces,” one Gartner analysis suggests.  If so, then edge computing revenue obviously grows.

The amount of data produced by things, the need to interact locally, and the need for real-time analysis will push computing towards the edge.

Such shifts in computing centralization and decentralization have happened before.

source: Gartner

In the early 1980s, client-server computing started to replace or augment mainframe computing, pushing computing toward the edge, the “edge” being the organization premises.

In one sense, the edge shifted a bit further in the mid-2000s, as smartphones gained widespread usage and mobile computing moved off the desktop.

At the same time, cloud computing also gained prevalence as most consumer mobile apps are cloud based. Generally speaking, the internet era has centralized computing at remote hyperscale data centers.

For at least some use cases, requiring ultra-low latency, edge computing is going to make more sense.

Yet it remains unclear which entities in the value chain will emerge as key suppliers of edge computing, given the numerous use cases. Autonomous vehicles are likely to require onboard computing. So might highly-immersive apps on gaming consoles, smartphones or other appliances using virtual reality or augmented reality.

Manufacturing plants might also require on-site, localized computing, if lots of sensor data controlling machinery is the use case.

Medical sensors might, in many cases, also require ultra-low latency possible only with local computing. Some military apps might be that category as well.

In fact, if computing happens “all over the place,” then defining what “edge” computing is will be difficult.

Enterprise IT, gaming vendors, telecom and cable operators, smartphone vendors, and consumer appliance vendors, app providers and cloud computing interests all have plausible stories to tell, in that regard.