What is Edge Computing and how it differs from Cloud computing?
In: Technology
You’re on the playground during recess and you’re discussing with your friend what 14 times 95 is (as cool kids do).
Your friend is pretty good with computers and could run back inside to Google what the answer is. (And this would still work even for harder problems!) That’s pretty much guaranteed to get the right answer, but it’s a pain for your friend to run all the way there, wait for a turn on the computer, Google it, and run all the way back.
But your friend is also good at mental math and can multiply it out in her head. She might not get the answer as fast as the computer would, but it’s still faster than running inside.
The second case is like edge computing — the “less powerful” devices immediately close to you, like your smartphone (or your friend, in my analogy), can do calculations and such directly so that you don’t need to send the questions to the more powerful cloud (the computer inside the school). Even though cloud computing might get the answer faster, you still save time overall for relatively simple questions because you’re not wasting time sending the data to the cloud (or running back inside). For real edge computing, you’re also just saving on data costs since you don’t need to contact the cloud at all.
But if a question is super complicated, like 1454 times 7395, then now your edge-computing friend might not be able to do it in her head, so she’ll still have to run inside and wait her turn to ask the cloud-computing school computer for the answer. That’s why cloud computing is still important.
Maybe for a medium-difficulty question, she doesn’t have to run all the way inside and can ask a nearby teacher instead, which still takes time but is faster than going all the way inside. That’s like another layer of edge compute.
—
The other comment gives a pretty good real-world example of these problems (video recognition), so I won’t repeat that here. 🙂
Edge means local. Netflix will have a few servers likely stored in your ISPs days center. On this server they cache or store movies they think will be popular. This way when you select the movie that’s new it will be more local to you.
Another example of edge computing is to put services local to the entity that needs them and then submit them centrally later likely to a cloud. Think of a fast food restaurant may have a server running inside of it to power all the needs of the store. This would be faster and more distributed than having it centralized for all the stores.
Cloud computing is using computers in the cloud to calculate things, so you send all your input data to the cloud, it calculates things, and you get the results back.
Edge computing is calculating a large portion of the necessary things already before uploading to the cloud, so that the amount of data (and thus both time and cost for data transfer) is minimized.
A popular example is video recognition. Without edge computing, you would stream the full video feed of a camera to the cloud, where it would be processed to identify things. With edge computing, the device with the camera would process what it sees, e.g. whether there is movement at all, and then crop the picture down to the relevant part, and then only transfer the really interesting parts to the cloud for further analysis.
So you are 5 and studying in Kindergarten. Now say you got to school to learn what edge computing is, but you see that one of your friends already learnt about it from the teacher and is on the way back. You meet him at the gate, and he explains it to you and you go back home instead of meeting your teacher 🙂
The gate is like an edge, and teacher/school is the core.
Latest Answers