Beyond the Cloud: Debunking common myths of edge computing

Far from a passive, near-perfunctory periphery of the network, the edge is a bustling locale for data analysis, management and storage. The migration from the centre of data’s gravity to the edge is transforming industries and opening up new market opportunities. In an October 2018 report, McKinsey & Company identified 107 distinct edge use cases, estimating the potential value of edge computing at $175 billion-$215 billion by 2025, and this is just the value for hardware companies. As per Gartner analyst Thomas J. Bittman, most companies are waking up to the reality that there is a need to expand beyond centralisation and the cloud, and towards location-based and distributed processing for low-latency and real-time processing. That said, some rather understandable misconceptions can cloud the edge. A look at the three most common myths and how they stack up against reality…

Myth: The edge will eat the cloud; Reality: The edge and the cloud will boost each other

With distributed computing on the rise, venture capitalists began to shift their priorities accordingly, with some issuing extreme forecasts. One such notable prediction was made in a 2017 talk given by enterprise investor Peter Levine, wherein he declared that with machine learning and IoT driving the shift to edge computing, the relevance of cloud will reduce in the near future. In the same year, Gartner’s Bittman issued a warning that the edge will eat the cloud and emphasised the shift towards location and distributed processing for low-latency and real-time processing.

A recent IDC study predicts that 30 per cent of the world’s data will need real-time processing by 2025. For instance, autonomous and connected vehicles are both intuitive edge use cases. If a connected or a self-driving car’s sensors detect that children are playing on the road while another vehicle is approaching at a fast pace, this information needs to be processed quickly. There is not enough time to send those insights back to the cloud for processing. Levine was right to point out that this life-critical data will need to be processed at end points, often via machine learning. However, this important information will still get stored in a centralised cloud, indicating that the cloud will become a learning centre of sorts to enable machine learning en masse, which requires aggregating data insights on the edge.

Thus, the edge will not overtake the cloud. Instead, it will prompt the cloud to extend its fabric to the edge. The hyperscale data centre model continues to work well for applications that benefit from centralisation such as large-scale archiving, content distribution, application storage and fast prototyping. Further, a specific kind of cloud deconsolidation is taking place concomitantly. According to “Data at the Edge”, a 2019 report published by Seagate with Vapor IO, companies like Vapor IO, EdgeConneX and DartPoints are turning to micro-modular data centres, also called edge data centres. They are small, regional, self-contained, cost-lowering, automated, micro-regional data centres at the edge of the network in locations such as parking lots and the base of cell towers. Designed to withstand environmental and security challenges at the periphery, these edge clusters have sufficient computing power to aggregate and process data separately from centralised data centres.

Myth: There is only one edge; Reality: There are many edges

There are a growing number of networks and, therefore, a growing number of outer network boundaries containing end points that run applications for users. These applications can be run in a barn, a field, a connected car, and a number of other locations. With time, the edges will become cloudified. Customisation will happen, but likely on the software layer.

Myth: Shrink the cloud, put it in a box and you have the edge; Reality: The edge is not a tiny cloud

Certain attributes of the cloud environment will need to be replicated across a variety of edges. These include equal network access and compatibility of an app developed in one edge network across different edge networks. This might make each edge look like a little cloud, but it is not the case. The edge infrastructure depends on the applications or use cases, which vary widely. These include utility regulation in smart cities, virtual reality scenarios, monitoring of ageing bridges, and robots making clothes in factories through virtual assistants. The edge will have no room and no time for certain types of data. Archival data or data needed for machine learning processes (data lakes, big clusters of data that teach machine learning algorithms) in the hyperscale data centre will be of no use at the edge.

Unlike the cloud, the edge is identified by a location and how near it is to the data source. Also, each edge focuses on solving a specific problem unlike the centralised, homogeneous, general-purpose data centre hub.