For many, this story would be a spin tale. As for people who do not understand cloud, fog computing can be a foggy affair. For years, we have been relying on hard disks and drives to drive our data consumption, but over the last half a decade technology has allowed storing massive amounts of data in the virtual space. Not only data companies have been able to run servers in the cloud, they have been doing big data analytics using cloud-as-a-service. That doesn’t mean companies have abandoned cloud altogether. Data servers are still up and running, and if anything companies are adding more might to them. Teradata, a data analytics giant, earlier this year announced a new iteration to both its data servers as well as its cloud technologies. More important, not only did it amalgamate these two technologies, it also allowed for a wider reach to its audiences using the best of both worlds.
Teradata and other companies’ experience shows that cloud is here to stay, but so are data servers. So, even if cloud becomes the next big thing, there are still solutions that would need a server. And, nobody describes this better than Marc Clark, director of cloud strategy at Teradata. “How do you move 3TB of data on the cloud?” he asks and then emphatically answers, “You don’t”.
But is Clark right? Is there no way to interact with technology except for clunky servers. Even Clark realises that technology is changing fast. When asked about edge and fog, he is not reluctant, rather he embraces it. He says “we will cross that bridge when we get there”. And when we will get there? The way technology and security needs are changing it would not be far away, and even Clark knows that.
The world is, no doubt, frantically moving towards cloud—even individuals have not been remiss in adopting it—but programmers have been working on competing technologies. And, edge computing is one of them. Edge is no new phenomenon. For those following grid computing—combining the strength of different computers to achieve a single goal—edge cannot be too difficult to understand. Edge, much like a grid, follows a similar pattern, where services, instead of interacting in the main frame of the cloud, interact at the edge of it. Being near the source of data, not every packet of data needs to be sent to the cloud and computing can be done between sensors at the edge of the network. Another iteration of this can be fog or mist, for medium and light weight usage respectively.
“It would be definitely beneficial for large data processing and also for low latency situations like some Internet of Things (IoT) cases. For example, a machine needs to stop when it reaches certain critical parameters say heat, speed, etc., to avoid damage. Round trip can be expensive and by the time an instruction is sent back to the machine, it may already be damaged. In such situations it is worthwhile to do data processing on the edge,” says Abhishek Bhattacharya, vice president, Sapient Global Markets, explaining how edge can be important.
And, IoT is what fog/edge is expected to drive. With smarter grids and smart everything, latency in data can cause problems. Imagine your autonomous car waiting 10 minutes to get the next decision from the cloud. Given how impatient technology has made us, not all of us would be willing or can afford to wait that long. This is where edge steps in. Instead of everything being transferred to the cloud, it will interact with the red light to determine whether to move or not. In other words, instead of passing on a message via a friend to another, now you will be able to do it yourself.
But IoT is not the only application edge is going to focus on. Being a closed loop network it can also ensure better security for certain applications; the difference, however, would be marginal. In any system where companies need speedier solutions, edge can be a beneficial tool. It is not to say cloud is completely out. Edge will do to cloud, what cloud has done to data servers. Not everybody would move to edge, but those who do will not abandon cloud either.
But as with any technology, there are problems to how edge gets implemented. Clouds could work well on 4G networks—larger transfers require broadband and fibre optic solutions. Similar is the case with edge, as with IoT devices, edge would require 5G and some countries are still far from that reality. But as sensors get better so would cloud, fog and mist. For economists, a cloudy or a misty outlook is a cause for concern, for technologists, it can be a foray into a new world of computing.