Cloud Computing Is Being Edged Out

September 8, 2017

Cloud computing is one of the most popular new forms of technology. It saves resources, offers greater flexibility, and according to 82% of the companies using the cloud, it saves money. However, according to, it’s officially passe.

The world of technology has always moved at a dizzying pace, with new software and devices constantly competing for the fleeting title of cutting edge. Still, the idea of cloud computing being left in the dust is striking, especially considering only two years ago Forbes was reporting that the cloud market was projected to reach $500 billion by the year 2020. Even more striking is the reason for the sudden pronouncement: the self-driving car.

It comes down to safety. While most accidents are the result of human error and negligence — the top three being distracted driving, driving drunk, and speeding — there are other times when an accident occurs through no fault of the driver. In those moments, a split second decision can be the difference between life and death.

But with the current cloud based processing, self-driving cars typically face a 100 milliseconds delay as the information travels from the large server farms, which are powerful enough to make such important decisions. But in that 100 milliseconds, it is often too late to prevent an accident.

“That problem from the frontier of technology is why many tech leaders foresee the need for a new ‘edge computing’ network—one that turns the logic of today’s cloud inside out,” Jeremy Hsu writes in his piece for

The problem is cloud data relies on massive data centers, which are all housed in one location. These buildings can be absolutely massive; the Des Moines Register reports that Apple Inc. plans to build a 400,000 square foot Data center in Iowa in order to keep up with the demands of their App Store and only their App Store.

Given the scale of these data centers, it would be impossible to build such a massive data center in crowded urban areas like New York City, Chicago, or Los Angeles. That means that the cities with the highest amount of traffic will be uniquely unsuited for the integration of self-driving cars.

The solution, according to start-ups like Manhattan-based Packet is to provide small clusters of servers scattered across a larger area. While the servers might not have the computing strength of a large data center like the one being built in Iowa, it has plenty of power that can greatly reduce the speed of computation to 15 milliseconds or less.

Major corporations are also taking an interest in these smaller server clusters, called edge networks. ATT announced plans in July to start incorporating edge computing into their 5G network in the hopes that they can reduce the delay into the single digits.

But while excitement continues to build around these edge networks, few people think they will truly replace more conventional methods of cloud computing.

As Hsu puts it, “Our era’s digital cathedrals are not vanishing anytime soon.”