The benefits of an increasing rise in edge data centres

Apr 03, 2020
author logo

Edge data centres (EDC) offer the advantages of conventional cloud computing services at cost-competitive prices coupled with lower latency periods due to their proximity to local end-users.


The need for EDCs is growing at the micro and macro level. At the micro-level, it’s being propelled by digital technology’s new applications that are transforming every aspect of how we work, live and travel. The cascading volumes of digital data and mobile services are triggering immense growth in the aggregate data demand.


At the macro level, the nation’s Digital India and Smart Cities’ missions, to name just two, are driving an unprecedented increase in annual growth for data services. Against this backdrop, the data centre industry needs to keep providing increasing data capacity as well as greater computing speeds simply to keep pace with the gargantuan demand. But centralized data centres and cloud computing support cannot overcome the challenge of latency for end-users in tier 2 and tier 3 cities.


Why Localization Wins


To overcome the latency barrier, localization of data centres is indispensable. EDCs overcome the latency hurdle by supporting computing and data storage needs as close to the end-users as possible. Although latency cannot be eliminated totally, it is majorly minimized. EDCs function on the principle of many smaller but well-distributed data centres providing a more viable solution as near the home location of consumers as possible.


In other words, burgeoning customer needs are driving the demand for EDCs. In fact, EDC locations are driven by diverse business imperatives, including application latency plus network capacity and costs. In India, working closely with the Centre and State Governments can benefit EDCs, given the concurrent and developing demand from Smart Cities and other institutional users. In these cases, EDC services can be delivered faster and seamlessly due to their geographical proximity. In such scenarios, one EDC could be connected to multiple EDCs.


Thereby, storing, processing and caching data and, thereafter, delivering the requisite information at the appropriate moment and the right place could be done speedily when needed. With EDCs, centralized data centres can then be side-lined because cloud features are available right at the site of operations, at cost-effective rates.


Of course, whereas EDCs can provide a series of services independently, they could use centralized data centres in supporting them for cloud services and analytics. But EDCs would still hold the edge in this symbiotic relationship since they function as co-located data setups where the data is processed and analysed closer to the client’s location. Thereby, cloud computing facilities could be offered to multiple companies.


Moreover, EDCs help customers limit operational costs. Whereas centralized data centres could be more than 100,000 square feet, EDCs can operate from relatively smaller areas via new designs and operating systems requiring lower investments. These co-located data centres benefit startups and traditional companies wherein proximity and cost containment are dominant elements in servicing demand. A major benefit of proximity lies in enabling almost real-time data analysis.


There is another way whereby edge computing lowers connectivity costs – eliminating the constant need for lightning-quick connectivity between conventional data centres and devices. Rather, super-quick connectivity is reserved for data that requires urgent processing. Meanwhile, data without urgency is serviced through alternate connectivity arrangements.


Demand and Distribution Dividends


It needs to be clarified, however, that conventional data centres will continue serving their purpose of processing all the core data where long-term retention is crucial. Of course, in all the arrangements, transparency is essential so that the customer clearly comprehends the cost differential and roles of EDCs and conventional DCs. Transparency can make it clear to customers why centralised cloud computing is not necessarily the best option for geographically-distributed applications.


A Gartner study reveals that the next generation of business innovation would be powered by the Internet of Things, artificial intelligence and 5G. All of these will propel a data production power surge where more than 7.5 billion connected devices will be used by enterprises in 2020 itself. Naturally, all these connected devices will not be located in metropolitan cities only. As a result, their diverse spread is incentivizing data centre infrastructure to be more distributed to reap the dividends of increasing demand.


Furthermore, the above conditions are creating a virtuous cycle and inducing changes in data centre infrastructure, which are gradually becoming more distributed, as stated earlier. As a consequence, more storage hubs are likely to emerge in smaller cities and other regional markets. In addition, micro data centres can be attached to sections of already-existing communications infrastructure, which includes telecom towers.


Much of these changes are already underway across India as the digital ecosystem keeps growing and influencing overall sectoral growth. But as EDCs are more widespread, it will call for a concomitant, widely-distributed surge in storage and security systems. Be they institutional or non-institutional users, they will demand similar levels of performance and security earlier offered by conventional data centres. Therefore, deliverables and efficacy offered by the centralised data centres would need to be replicated by EDCs for their lower-latency services model to enjoy the same success as the former.


Finally, for both customers and EDCs, such storage and security solutions will have to be determined at the beginning of the planning processes before the new models are poised for rollout. Thanks to the nationwide drive towards a smart digital future, sooner or later, India’s ICT industry will witness the mass emergence of EDCs across its landscape, benefitting both end-users and deployers.

Read more