It’s funny really, Cloud took the whole world by storm and no one was prepared… then it evolved becoming something new… then something more. Now, we see the exact same thing again; no one understands it nor is prepared for it, they just know that a new age is dawning.
People are seeing Edge to replace Cloud while not even understanding that one’s not the replacement of the other, but a fundamental complementation of the other; the other side of the Coin. There is even a middle ground where the two meet in perfect harmony….
Today we fix that, today we Demystify Edge Computing vs. Cloud Computing.
“Edge computing is likely to work in tandem with cloud computing — not replacing it.”
Cloud is basically using a remote server for the handling of data; it’s a centralized, physical construct that can be accessed form anywhere at any time. This gives it great versatility, and utility under the right scenarios, but does have some limitations as per construct; it physically exists somewhere.
In Cloud Computing data is sent to the cloud server, processed there then pulled back by the customer as needed; the draw back is that, in most cases, the data is located far from point of origin and the desired recipient. This means that there is travel time and this creates issues in some cases…
Since it’s stored in the cloud and these units are located off-shore, issues of latency, freshness, and accuracy of said data becomes problematic. Especially, when promptness of information transit and accuracy of that data in real-time are points of concern.
However, this is no issue for things like images, videos, music and the like where that isn’t a factor; making this ideal for these mediums.
Edge is an ideologic construct of moving the processes of a function towards the “edge” of the network and the source of the information input; i.e. the user and the device.
Thus, Edge Computing using the same devices to gather, process, and deliver information; it’s even used as the storage point most of the time.
This ideologic construct is the solution to the drawbacks of Cloud where those points of concern meet the limiting factors of Cloud. This however is NOT the same as cloud, it is the solution to clouds problems…
The purpose of this is to alleviate the problems of Cloud Computing, but it has a fundamental drawback, Storage Space.
As edge functions on the gathering device its limited by the hardware of the input devices for both processing and storage capacity; a problem Cloud itself doesn’t have.
Think of it like this, each has their place where it shines and would be terrible if in the others situation. Edge Computing is great for automated car functions like Honda Sensing, while Cloud is great for things like Music Streaming.
Now imagine for a second how bad these would be running on the others system. But what if you could cut away each one’s flaws without giving up their strengths?
Wouldn’t that be ideal?
Nah, that’s just a pipe dream… or is it?
Guess what… it exists and it’s what’s causing all this confusion over edge vs. cloud, it’s CLOUD EDGE.
Edge has got its place to shine and so does cloud, but what if you needed both… well this has got you covered.
It takes Edge’s ideologic construct of moving the processes of a function towards the “edge” of the network and the source of the information input and applies it to the physical construct of cloud servers.
Now this may seem weird but all it really is, is taking the centralized servers for cloud and decentralizing them and spreading them locally in smaller batch nodes. That said, it goes a little farther then either process can go on its own.
The edge devices do the compilation of data it sucks up there and feeds the results back to user, while storing the data server side to reduce the load on the devices.
This data then can have the computation results compiled and applied to machine learning. The data then allows for predictive analytics of the results and facilitates deriving patterns.
This derived data is fed back to the device and when a repeat or similar query is made by the user or similar data is collected, it can use the derived data to augment its process of output for said query.
Then it’s saved in the local node and becomes relevant only when applied to the local area, thus only needs to be stored locally rather than in a centralized repository. This gives for fast, smart and accurate results with history behind the decisions and speed close to that of a stand-alone Edge ideology.
While this sounds complicated it can be summed up in a single, simple application example; iPhone Maps app. Edge for active guidance/tracking, detours, and accident awareness, cloud for predicting the best route when you input an address.
It also uses other users edge device capabilities to augment your results based on their active conditions and actions. All the strengths of edge with the benefits of cloud.
Sure, it gives up a little speed to be able to use this cloud function, but it functions far better then if it was using either system independently.