Edge computing is no longer a hype. It is real and is gaining traction globally. The demand for low latency and real-time data processing is making businesses increasingly aware of the need to expand beyond traditional cloud, towards location-based and distributed data processing. A recent IDC study predicts that 30 per cent of the world’s data will be processed in real time by 2025. An increasing number of companies are realising its benefits, and are in the process of adding Edge elements to their centralised core networks.
Edge computing brings networks, data storage and computing closer to users. This geographical proximity to users promises significantly lower latencies, thus eliminating lags. The aim is to ensure that data from sensors and other devices gets aggregated and analysed as close as possible to its source, and the content is stored close to its consumers. Of course, this is a striking departure from most of the existing computing designs, wherein queries are processed in data centres located thousands of kilometres away from the source of data or the point of consumption.
McKinsey & Company, in its report released in October 2018, identified 107 distinct Edge use cases, estimating the potential market value of Edge computing at $175 billion-$215 billion by 2025 (the value for hardware companies). The key use cases of Edge computing are augmented and virtual reality (AR/VR), industrial applications, autonomous driving/connected cars, and other advanced applications that are enabled through a low-latency, high bandwidth and local processing platform rather than centralised cloud. According to B.S. Teh, senior vice-president, global sales and sales operations, Seagate Technology, if a connected or a self-driving car detects that 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. This will require aggregating and analysing data directly on the Edge.
Further, Edge computing can be extremely useful in the healthcare industry as it can help medical professionals gain real-time access to patient data, thereby improving diagnosis and patient health monitoring. As for industrial manufacturing, Edge computing can help manufacturers gather data and gain insights for predictive maintenance and energy efficiency.
Edge computing is a perfect example of convergence of networks and IT. The concept complements other technology evolutions being witnessed in the telecom space, such as virtualised system deployments, advancements in distributed network architectures and, most importantly, the introduction of 5G.
Gaining an Edge: Telcos’ evolving role
With data taking centre stage, the commercialisation of telco Edge has started. With the growing consumption of video content and the surge in mobile internet services, consumers today are demanding ultra high-speed, low-latency and seamless digital experience on the go. The delivery of such an experience is only possible if the compute power of networks resides as close to the consumer as possible. Distributed Edge computing is critical to ensure a true digital experience. Operators with access to distributed infrastructure are best suited to scale up the Edge computing concept. They have a presence in remote locations and thus, in the long run, may even gain an Edge over cloud giants such as Amazon, Google and Microsoft. With the emergence of 5G, telco Edge deployments are likely to emerge as the best platforms to deliver services close to end users and enterprise facilities.
Currently, the global telecom market is focusing on the deployment of Edge computing through several initiatives, including multi-access Edge computing, open Edge computing, and independent vendor activities. Indian telcos too have made announcements around Edge computing (see Box). To leverage the Edge opportunity, telcos need to understand the requirements of the industries that they serve and the benefits Edge computing can bring to these industries. Further, they must identify enterprise verticals that need to be prioritised and the ideal location for the Edge servers for these verticals.
Edge and 5G: A strong combo
It is an accepted fact that a cost-efficient service delivery for 5G will not be feasible without the deployment of some form of Edge data computing. 5G is not going to be a mere connectivity service like the previous technology generations of 3G and 4G. Most of the 5G deployments will be localised and will require an element of local computing.
Edge computing can ensure that 5G latency is so low that it becomes indistinguishable from reality. Broadly, the use cases of 5G Edge computing will be premises based and network based. Premises-based refers to cellular networks that connect to on-site servers that have cloud software plugged into them. The technology can be used to enable remote patient care, and leverage AR and artificial intelligence (AI) for student training. As for network Edge compute, it allows businesses that have more distributed environments to use the mobile network or the fibre network, along with cloud. Network edge computing uses metro-based servers. Globally, companies in the healthcare, manufacturing, corporate training and drone detection industries have already started adopting 5G Edge computing solutions.
With the commercialisation of 5G and AI-driven IoT, Edge computing environments are set to create a bigger disruption in the telecom space than cloud computing.
Micro data centres: An opportunity for tower companies
Micro data centres, also known as Edge data centres, are small, regional, cost-effective, automated, micro-modular data sites that can power high-speed computing. Globally, telecom tower sites are emerging as the perfect infrastructure for hosting these micro data centres. Telecom tower sites have a steady power supply, ready access to fibre backhaul connectivity, the requisite real estate, and they are located right at the network Edge. Often, there is enough space at the base of towers, which is an ideal place to install these data centres. Fibre connectivity can help connect the Edge sites to the core/centralised data centres.
Towercos across the world are currently engaged in discussions with players from various industries, which may ultimately become Edge computing tenants. In India, the telecom tower count stands at well above 500,000 sites, and each one of these can be transformed into micro data centres. This presents a unique opportunity for towercos, which can add new revenue streams and customers by leveraging their assets. By deploying Edge data centres on their sites, towercos can onboard customers with stringent latency requirements such as content delivery network providers and cloud providers, which support Edge-specific applications.
India’s data centre footprint may well exceed 100 million square feet by 2060, comprising 5,000 Edge data centres. These micro data centres are likely to come up at almost all major metros across the country.
The addition of Edge computing to core networks will pose several challenges. The new set up will be more complex and the regular data centre operators might not be equipped to deal with. The likely increase in the number of small data centres close to end users would still require the same performance and security as experienced in centralised data centres. Enterprises will have to undertake data centre infrastructure management to ensure that the distributed IT facilities are secure.
As per industry estimates, about 75 per cent of data will have to be processed at the Edge by 2022. This trend of Edge network deployment at data centres will gain momentum with the introduction of 5G networks and IoT mainstreaming. This, however, does not signal the end of hyperscale data centres’ era as these will continue to be used for applications such as storage, content distribution and large-scale archiving. The Edge-driven systems will work alongside the current data centre model.
Going forward, the adoption of Edge networks will continue to grow with the increasing proliferation of IoT-enabled devices. According to IDC, by 2023, around 45 per cent of IoT-generated data will be stored, processed, analysed and leveraged close to, or at the Edge of the network, and over 6 billion devices will be connected to Edge computing solutions. s
By Akanksha Mahajan Marwah