The rise of connected ecosystem and new experiences powered by technologies such as AR/VR are putting increased pressure on limited bandwidth in centralized computing architectures. Moreover, latency-sensitive applications such as virtual assistants and autonomous cars cannot enjoy widespread adoption if the present centralized computing architecture is not able to ensure real-time decision- making. Hence, there is a need to move computation closer to the point of data generation/consumption to reduce bandwidth pressure and power uses that have tight real-time decision-making requirements. This concept of moving storage, computation, and networking away from centralized computing to the proximity of the network is referred to as edge computing. It encapsulates a set of technologies that enable distributed computing and embeds intelligence at the proximity of the network.
In edge computing, it is important to understand what the proximity, or the edge, of the network really implies. Is it a device, a location, or simply a concept? Moreover, how is edge computing different from cloud computing? Is it a new computing paradigm that will mean the end of cloud computing or does it exist in association with it?
This paper demystifies edge computing. In it, we:
Define edge computing and its key characteristics
Establish the symbiotic relationship between edge and cloud computing
Introduce a decision-making framework for edge computing adoption
Provide an overview of ecosystem enablers for edge computing
Explore key considerations for enterprises seeking to adopt edge computing
