Is edge computing a real solution to data loss prevention?

When Pokémon Go came out, it ran on one of the large cloud players. That’s one app over 4G with millions of people all over the world suddenly engaging with it. Because of the surge on cloud they needed to find 40K physical servers in 48 hours to handle the demand from people locally initiating this. And this goes to seconds: They could wait a few seconds as they're walking around trying to find a Pokemon with their phone. So it still went back to cloud. But the burst of one application over a 4G network speed, with a latency of anywhere from 50 to 300 milliseconds, required 40K servers-worth of actual performance increase to supply that.

Now we have 5G—any developer can create Pokémon Go again. There will be a forcing function and a surge. It really puts into perspective how many more will go out there. That's why this data tsunami is just massive. There are some very interesting things that will come up so we've got to keep our eye on this side.

Thinking of all that data loss, I can't help but be reminded of the early days of Standard Oil, when they used to dump gasoline down the river because there was no use for it. I feel we are on the cusp of something as big as that with edge computing.

We are producing insane amounts of data. But if we don't know what we can do with the data, data is useless. That's why you will see a lot of data being lost early on. But as people start understanding the power of data, they will start capturing it from the edge.

The question is whether to do asynchronous or synchronous capturing and transfer. Those depend on the quality of bandwidth and of the compute power used to transfer that data. A lot of companies are looking at asynchronous transfer. The edge device or edge paradigm has enough to store it for a few hours or a day and then it can transfer the data one batch at a time, to make sure it is not overwhelmed.