This is a prediction that not many people tend to make. But, the fact remains, the cloud computing market, as it is currently known, will become obsolete in nothing more than a few years.
The Rise of IoT
The person who is predicting these things is Peter Levine, a partner at Andreessen Horowitz. He says that the Internet of Things devices have an increased computing power which has been combined with the increasingly accurate machine learning technology, and it is going to eventually replace infrastructure as a public cloud market service.
The cloud, as it is known currently, is basically a centralized model in terms of computing according to Levine. Information is uploaded to the cloud where it is then processed before being stored. There are a lot of applications which live inside the cloud, and complete data centers are also getting migrated to it.
Levine has said that we are already trying to see any signs of the Internet of Things devices coming in place of some of the cloud’s computing power. There are smart cars, robots, drones, machines, and appliances. All of the devices mentioned here collect real-time data. The networks do have to deal with latency, and there’s a lot of information too.
This is why there is not enough time for the information to go back to the cloud so that it can get processed. He also argued that the edge of this network is going to be forced to get more sophisticated. He added that the shift was going to decimate cloud computing as it is known today.
For example, let’s take a look at self-driving cars. They have to be able to identify pedestrians or stop signs and then act on the information they gather instantaneously. They will not be able to wait long enough for network connections with the cloud in order to figure out what to do.
This new world isn’t going to completely rid us of the need for centralized clouds according to Levine. He says that the cloud is still going to be in the same place where the information is getting offloaded to, where it’s being stored for longer periods and where the machine-learning algorithms can get access to the vast amounts of data that they would need in order to get smarter.
The idea here where edge computing is going to become more powerful than everything else isn’t something entirely new. Cisco is the one who should be given the credit for coming up with this term in the form of Fog Computing. This basically is the idea of acting and analyzing time-sensitive data at the edge of a network.
Levine says that all of this is quite like a back to the future moment. Computing started back when there was a centralized model which focused on a mainframe. After this came the world of distributed client-servers. It was the cloud which swung the computing world back to using a centralized platform. Now this edge intelligence is going to swing the world right back to a distributed system.