Cloud, Edge, Fog — what do these three words mean when they refer to computing, and why do they often appear together?
The world of computer processing is rapidly changing. Wireless speeds and data aggregation processes are evolving at top speed. Those developments become easier to comprehend when we clarify Cloud, Edge, and Fog computing.
Cloud computing or cloud processing happens fairly often in a home technology environment. In this case, the “cloud” is simply the web that exists outside of your local network. Currently, every single voice speaker uses cloud computing to analyze what you have said after you have given a device the activation word or phrase. The processing power to go through all of the possible scenarios is tremendous, hence the need for interaction with the larger internet.
And there’s the problem: When you ask about, say, a date in history, the data that can be tracked in that user/web interface is, on the surface, fairly innocuous. But when it comes to privacy, many people are starting to get nervous about all their requests being out in the cloud for anyone to access.
As other analytic software solutions come into the home, they are also generally going “into the cloud.” Take security: Video and audio analytics that monitor the home require powerful processing that only the cloud can currently handle.
As processing power and silicon prices drop, the ability to put more robust power on devices rises. The voice speaker manufacturers are already talking about bringing the processing into the local realm for performance and security reasons. This is called edge computing.
Fog bridges the gap between edge and the cloud.
Generally, edge computing will be used for a very specific task or role as the cloud is still utilized for data aggregation, storage, and so on. The growing number of sensors and the need for instant processing results is putting more pressure on the adoption of edge computing. Driverless cars will be utilizing some of the fastest and most robust edge computing on the market when they roll out. The ability to make instantaneous decisions based on AI and machine learning without the inherent latency of a cloud-based system is key when it comes to the functions and decisions an autonomous vehicle will need to handle.
With enterprise-grade processing coming to home electronics, we will start to see a blend of cloud, edge, and fog computing. Fog bridges the gap between edge and the cloud. Generally, a centralized processor will live on the same network as edge devices, giving them a second level of processing before having to go to the cloud. In a residence, these would generally be our whole-home control processors. Edge processing gathers data from all the subsystems, managing the data and deciding when to go to the cloud for additional resources. Smart hubs also act as fog devices since the cost of placing all of the processing power into, say, a small light bulb would be prohibitive. Edge and fog also allow for a much tighter level of security, keeping your data safe and secure internal to your LAN.
While technology continues to double its computing capacity every 18 months (Moore’s Law), processing power and sharing the load of that processing across the local edge, midway fog and overseeing cloud is going to be key. As our homes become more complex and packed full of sensors and connected devices, these terms will be become common within the custom integration industry.
NOTE: This Emerging Trends piece is brought to you by CEDIA’s Technology Advisory Council and Technology Application & Innovation department.