The cloud is dead.
OK, maybe not dead. But definitely going out of style.
That, in effect, has been the subject of a spate of recent articles, at least a couple of which cite Peter Levine of Andreessen Horowitz as a key source.
The same week those pieces came out, Gartner (News - Alert) analyst Thomas Bittman authored a piece of a similar flavor titled “The Edge Will Eat the Cloud.” (This title is, of course, a reference to Mark Andreessen’s oft-cited 2011 Wall Street Journal article.)
A more recent Wall Street Journal piece, this one from March 7, 2017, quotes Levine commenting: “I have this theory that cloud computing goes away in the not-too-distant future.”
Levine has told various media outlets that cloud computing, in his mind, is analogous to mainframe computing. Both are centralized.
But the Internet of Things, Levine says, will call for decentralization – or computing at the edge.
No doubt. The IoT will absolutely require more edge computing.
But this should come as news to no one.
And the ultra reliable low latency communications that 5G is expected to help enable, we’ve been told again and again, will help cellular service providers support applications like the connected car and remote surgery that require mobile edge computing for faster decision making.
I mentioned that in a story about 5G that ran in the first quarter 2017 issue of INTERNET TELEPHONY magazine. But Yun Chao Hu, Milan Patel, Dario Sabella, Nirit Sprecher, and Valerie Young authored a European Telecommunications Standards Institute white paper on mobile edge computing way back in September of 2015.
Access Network and in close proximity to mobile subscribers,” the 2015 paper explained. “The aim is to reduce latency, ensure highly efficient network operation and service delivery, and offer an improved user experience.”
As the paper explained, edge computing doesn’t necessarily signal the end of the cloud. Instead, it can be viewed as an expansion of the cloud.
Bittman, however, argues that “it’s more likely the edge will carry its own weight, driven more by consumers and consumer experiences than enterprises.” Edge computing, he adds, entails a different kind of workload and “will create some serious winners and losers, both in terms of vendors, and businesses.”
But I would argue that the idea that cloud computing is on its death bed is more than a little overstated. And it seems to me that the notion that expanding some computing to the edge will have earthshattering effects on cloud service providers is itself a shaky proposition.
Makes for a great headline though.
Instead, I think what we’ll see is the continuation of a healthy cloud, paired with edge computing for select applications.
One more note: Some folks have tied the discussion about edge computing to the recent Amazon Web Services (News - Alert) outage. For example, this K4Connect blog says edge/fog architectures could be designed not only to eliminate latency, but to operate even when the “internet” is down.
That would be great.
But this piece also goes on to note that (although the cloud, like all computing platforms I might add, is not foolproof) “in some instances – especially if the task requires significant computing power and there is no need for ‘immediacy’ – the cloud is the more appropriate solution.”
Edited by Stefania Viscusi