As the Internet of Things (IoT) grows more and more commonplace in our current IT setups, and “smart” devices will invariably become indispensable. An effect of this phenomenal growth in IoT is the overwhelming increase in the amount of data we’ll generate. According to a report by IDC, the amount of data generated annually is projected to increase from 40 zettabytes in 2019 to 175 zettabytes in 2025. This data deluge would mean new challenges – because although cloud computing remains the preferred option for enterprises, its scalability seeming infinite, it does have its limits. Latency is increasingly going to become a closely watched metric as the amount of data we produce increases, and IoT applications like autonomous cars would certainly require real-time processing. Security would also be a concern. A solution to all these challenges comes in the form of edge computing.
Edge computing refers to the storage and processing of data at the “edge” of our networks, i.e., where the data is produced. Rather than depending on a centralized and far-off cloud to process the data, workloads are executed locally, ‘on-device’. Essentially, edge computing would serve to decentralize the handling of data. This is said to have a host of advantages, especially when it comes to real-time data processing requirements as in autonomous cars and other IoT use cases. Exactly what impact will this have on global data connectivity? Let’s find out.
New Technologies For Connectivity
Several new-age IoT applications require data to be processed in real-time, and as we continue to generate increasingly large amounts of data, latency will position itself as a critical metric. Not only will this necessitate innovation in IoT devices and their on-board processing capabilities, it will also spur demand for mainstream, reliable 5G across regions. 5G will form the bedrock of the instant connectivity capabilities edge computing that promises, culminating in a symbiotic relationship that should assist all sectors.
Connectivity costs should also see significant reductions, as devices will no longer need to transmit every bit of data to data centres. Instead, edge will be removing the need for super-fast, high-bandwidth connectivity between IoT devices and traditional data centres. With over 75 billion active connected devices predicted by 2025, network congestion may be an issue resolved well in advance with the combination of 5G and edge clusters.
The edge would also have an influence on privacy and security aspects. With edge devices handling localised data, the damage potential from intrusions and threats may be minimised. Moreover, in case of a security breach, the decentralised nature of the edge will also make it easier to isolate such breaches, with security protocols in place for affected sections without causing damage to the entire network.
The Impact of Edge Computing on Data Centres
Edge computing would also induce a major change in how data centres operate. A large number of micro-data centres and edge-driven systems would be set up, supplementing existing data centres.
-
- Consumer internet traffic may be the biggest beneficiary here. As per Cisco’s forecasts – by 2022, video traffic will account for 82% of all business and consumer IP traffic. Edge clusters would quickly fill the need here, functioning as mini CDNs enabling quicker, and more reliable streaming for this large chunk of internet traffic.
- Edge computing’s impact on data connectivity will also translate into direct, measurable savings in cost. According to Schneider Electric’s estimates, a business that incurs about $7 million in capital expenses for running a conventional 1 MW data centre may end up spending nearly half as much – $4 million – if it chose to build five 200 KW smaller centres. For developing markets, this will be a boon.
Of course, smaller edge clusters are not a complete replacement for traditional data centres. Critical data would remain that may require long-term retention on the cloud, and the existing data centre model would remain useful for a wide range of applications. However, there is no denying that the rise of edge computing will bring about a change in how data centres operate. Several small data centres will be built closer to customers rather than in remote regions, and our data centre infrastructure will become significantly more distributed. Small data centres are also likely to be constructed on existing communications infrastructure, like in telecom towers.
Edge computing is likely to go a long way in helping increase speed and handling security threats when it comes to real-time processing of data collected from IoT devices. The predicted deluge of data is unlikely to be managed without edge computing infrastructure, along with some major changes in the data centre landscape. And, although the newer infrastructure will not entirely replace the traditional data centres, data centres will certainly need to adapt to find a place in a new data landscape of the future.
Sunando comes from an IT services background, and has over 23 years experience with companies like Orange Business Services, Spectranet, Sify and NTT-NetMagic wherein he was responsible for Business Strategy, Sales, Alliances, Product Development and Pre-sales functions. At IndiQus, he drives the entire business strategy, and also leads the HR and Finance teams. Sunando did his SMP from IIM Calcutta.