High latency caused by data centres being far away from end users is no longer acceptable as distance has become a digital business killer.
Digital leading enterprises now favour a distributed model of co-location and private interconnection to rich ecosystems, offering the lowest latency to both public and private cloud providers whilst ensuring high levels of security and reliability.
The digital economy continues to drive an increase in worldwide interconnection bandwidth with a forecast compound annual growth rate (CAGR) of 44 per cent by 2024 (4x the 2020 total), reaching 21,485+ terabits per second (Tbps), or 85 zettabytes (ZB) per year.
“With the digital economy only becoming more pervasive, companies are accelerating their shift to a digital-first approach and looking to capitalise on the market opportunities ahead by distributing their core infrastructure, extending to the edge to be closer to users, and interacting with business ecosystems in strategic locations,” said Craig Matsumoto, senior analyst, Data centre Networking, author of new Equinix report ‘Architecting Hybrid IT and Edge in a Distributed, Digital World’.
“Organisations are increasing their reliance on cloud-based technologies and automation of IT infrastructure operations with a desire to consume infrastructure and IT ‘as-a-service’, delivered with cloud-like consumption models. Enterprises are learning that they cannot do everything themselves.”
Enterprises are moving some workloads out of the cloud and gravitating toward a hybrid IT model. But backhauling all edge traffic to a centralised data centre, as in the old model, is not sustainable. In addition, real-time user experiences, edge analytics, workforce enablement and growing customer demands require distributing work and data to more locations connected to a wider variety of business partners. Edge computing placed in multi-tenant data centres with direct access to ecosystem partners can satisfy these requirements.
Latency looms large, especially for high-performance edge applications, IoT and 5G use cases. The distance to critical resources and users is becoming a crucial factor, one that enterprises can address by taking advantage of edge computing in ecosystem-dense co-location facilities.
Edge workloads are becoming increasingly critical to enterprises across all industries. The trends to support the workforce (think SD-WAN and SASE for now-remote employees), offer real-time analytics and transform the customer experience continue to be accelerated by the compounding effects of IoT, AI and 5G.
Bringing services closer to users while simultaneously increasing performance to support real-time needs and better user experience requires a completely new model. Strategic locations should be identified based on their proximity to large populations of users and the ability to co-locate the ecosystem partners required across the supply chain. Workload functions should be analysed based on whether they can be virtualised at scale and what benefits they can gain from residing at different ‘edge’ locations.
“As enterprises increase their dependence on clouds, SaaS providers and other outside entities, they can also rethink their networking profiles to regain control of connectivity,” Matsumoto continued. “A multi-tenant data centre already hosts many commonly required destinations and can transparently provide connectivity over private links. For hybrid IT and edge computing, this essentially removes the distance to business-critical ecosystems. A cloud on-ramp or a supply chain partner can be just a cross connect away.
“The advent of software-programmable interconnection extends reach, giving a co-location site access to many destinations around a fabric of data centres. This creates economies of aggregation, as the enterprise can enjoy the network effect of being surrounded by connectivity options, dense ecosystems and services that can now be consumed in a more cloud-like, on-demand manner.”
