Edge computing refers to a distributed IT infrastructure where the processing power is decentralised, allowing data to be collected, organised, analysed and processed locally and near to its source.
Think of a usual setup where devices are connected to a server whether in an on-prem data centre or the cloud. The data generated needs to be transferred through the network to these centralised locations for processing, which may suffer from latency, data loss and even downtime.
With edge computing, as the name suggests, the computing takes place on the edge of the network or at the end devices without having to transmit it to a data centre or cloud. Since edge devices are located near where the data is created, this results in much faster processing and analysis of data – enabling analytics and insights to take place in real-time – as well as a higher degree of computational efficiency.
Edge computing addresses some of the drawbacks of centralised computing infrastructure, such as bandwidth limitations, latency and security due to the fact that processes are carried out near the source of data. It also reduces data redundancy, because all the generated data does not have to be replicated to a centralised location. In such cases, edge computing can help organisations save costs in terms of reducing bandwidth usage and the amount of redundant data. They can even limit the amount of private or sensitive information being sent across a network.
Edge computing is a perfect fit for IoT applications because, with the exponential growth of data as well as IoT devices themselves, it isn’t feasible for organisations to send tremendous amounts of data from thousands or millions of devices to a centralised server or cloud. Edge devices can also function reliably in remote areas where internet connectivity is often a problem.
By avoiding network latency, edge processing has become an important element in powering IoT use cases that require precise operations or sub-millisecond response times, such as in the case of autonomous vehicles.
Edge computing is certainly a powerful resource and strategy but it should not be seen as a total replacement for cloud computing. When used in complement with the cloud, edge computing can help optimise the flow of enterprise data, maximise operating costs and enable reliable and secure local data processing at very low latencies.
