Information technology is a world where imposing names often mask how simple technology can be. Cloud computing is basically a business model where you lease applications on demand through the Internet. Earlier, this was known as centralized computing.
These were centralized computers which did the number crunching, and you used “dumb terminals” to connect to them without having any computing power. You could just see what was happening on the mainframe.
Moore’s Law has made computing prices fall exponentially with time, and we then in a time of decentralized computing where everyone had a computer which can do sizeable tasks. They used LAN for connecting to servers which then did the heavy lifting.
Then came e-commerce where we moved to data centers which were large server farms. From there, the best way forward was building huge data centers and called it the cloud. Then it was decided that everything needed to be connected to the cloud because that’s how things work today. This means that everything links to the cloud and the big data accumulated increases the efficiency of society. One problem faced here was a shortage of bandwidth.
Solving this problem was quite simple. This is where fog computing came into the picture. It turns out if you move computers closer to the devices they are communicating with, they get better bandwidth. Anyone who uses Wi-Fi on smartphones will know how this works. Fog computing involves moving some computing power closer to devices, so the cloud doesn’t need to be consulted for everything.
There are some time-sensitive computational decisions which can be made by intermediary devices which then aggregate all the data learned and sent to the cloud. This is sort of like decentralized computing. But, here we are talking about fog nodes.
Fog nodes sit near devices and fog aggregation nodes, a concept which is equivalent to the way in which workstations are communicated with servers earlier.
Edge computing is basically another name for fog computing. Computers which connects with all the devices in the cloud are called fog computing or edge computing. They are the same.
But these are overly simplified concepts which rehash ideas from the past. The real opportunity is related to configuring nodes and optimizing performance. The main difference between devices communicating with nodes and the cloud is bi-directional communication between nodes taking milliseconds while it takes minutes on the cloud.
Hosting analytics, performance processing, and heterogeneous applications closer to physical centers and control systems, can help enable edge intelligence.
They will help you move portions from cloud-based applications closer to devices which use them. It isn’t easy to figure out what software tasks to remove from the cloud, but the growth of bandwidth-consuming devices may force you to take a different approach.
For investors, there are not any opportunities right now to invest in pure-play companies for edge or fog computing. Cisco may be making a move, but there are a few other startups which may be considering IPO’s.