Edge computing might give us faster machines

edge computing

It’s 2017 and we are still able to work significantly faster than our computers, all too frequently having to stare at our screens, waiting for our machines to catch up with our thoughts.

The time between double-clicking on an icon, for example, and the moment when the document or application finally opens is just too long for our trains of thoughts to stay on track – more likely to get bored and run out of steam.

Even more frustratingly, the time delay is not long enough to go and get another cup of coffee.

What to do? Faster machine? Faster network? Probably neither. They may be slightly faster, but an increasing number of experts say the answer is edge computing.

Edge computing just means that more of the processing is done closer to the user, perhaps on business premises, or even on the machine itself, rather than at some remote data centre.

It’s probably not a cure for all ills, but anything but this constant stop-start train of thought – it’s enough to demotivate you and discourage you from starting projects that you know you would actually enjoy if you didn’t have to deal with the buffering or whatever it is that slows down the machine you’re working on.

Yes, a better, more expensive computer may improve your experience, but having worked on probably hundreds of machines over many decades, you’ve probably found that no machine yet has been faster than its expert human operators.

And these days, what with most people’s work almost always involving switching between many applications and websites and web-based apps and communications systems and so on… machines are being asked to do a whole lot more than they ever were before.

And that’s not even including the demands of big data, virtual and augmented reality, connected and autonomous cars, and all the other technologies in the pipeline, particularly those that depend on the internet of things.

But, according to Monica Paolini, principal analyst at Senza Fili – and many others – edge computing could be the solution.

“The edge cloud and edge computing will play a vital role in application and content delivery across both fixed and mobile networks around the world,” says Paolini. “This distributed layer of shared compute infrastructure in last mile networks, at home and at work, is the final element in the new architecture for application delivery.”

There are five essential attributes of a service provider of edge computing, according to Qwilt, which describes itself as an “open edge cloud” service. These attributes are:

  1. common compute and storage standards;
  2. massively distributed and shared;
  3. accessible to publishers via an Open API;
  4. matches resources to application and content delivery demands; and
  5. extends, augments and complements centralized cloud.

Qwilt points to research by Oculus, Nokia Bell Labs and Valve all saying that “extremely low latency is mandatory for the success of AR and VR”, but even moderately low latency wouldn’t be bad for everyday, basic applications either.

Qwilt estimates that the maximum tolerable network delay for some applications is 10 milliseconds or less. That sounds reasonable, and it seems to be what the punters want as well, according to the company.

“We are in a race to enable our open edge cloud to keep up with growing consumer demand for streaming media,” said Alon Maor, CEO of Qwilt. “By partnering with service providers, we can satisfy publisher requirements for edge cloud resources that deliver the requisite scale and quality.”

In another article, this one on Forbes.com, digital business architect Janikaram MSV suggests that edge computing is inevitable now that enterprises have “crossed the cloud chasm”.

Some business sectors – such as finance and the public sector – have been more sceptical, but are now increasingly adopting cloud, and the next big thing will be edge computing, according to Janikaram.

He says edge computing is all set to become the most preferred architecture for running data-driven, intelligent applications.

“The affordability of compute and storage combined with the rise of machine learning will drive the adoption of edge computing,” he says. “It’s not just IoT, even traditional business applications will start to take advantage of this architecture.”

We need faster machines and networks and edge computing might have the answer Click To Tweet