Edge computing has become an increasingly hot topic in the technology and telecoms world. However, the ecosystem is still evolving and a clear market hierarchy of leading players has yet to fully emerge. Edge is supposed to help deliver workloads at lower latency and with improved data security. This should be achieved by moving compute and storage resources closer to where end user devices are collecting and processing data and running applications. From the early days of 5G development, telecom operators talked about the possibilities of edge paired with 5G. Low latency is one of the key attributes of 5G networks. Many operators see an opportunity to include edge capabilities as part of their 5G build outs.
Edge ambitions from operators
SK Telecom (SKT) in South Korea was one of the earliest to launch 5G services. It has been a leading cheerleader for edge computing within the operator community. SKT launched a trial of 5G with mobile edge computing as early as December 2018 for an autoparts manufacturer. Since then SKT has rolled out mobile edge computing capabilities to at least a dozen of its own data centres. These are located in its existing offices across South Korea. Korea Telecom, the second leading mobile operator in South Korea has also launched mobile edge centres in eight locations.
Telcos are not the only businesses looking to capitalise on the potential. Currently, there a host of players looking to grow the value of the market and grab as large a share as possible. At the forefront of this movement are the global hyperscale cloud providers such as Amazon Web Services (AWS), Microsoft Azure, Google Cloud and Alibaba. All have invested in developing IoT edge computing capabilities that align with their existing cloud and IoT ecosystems.
The edge functions like an extension of the cloud closer to end user devices. Therefore it makes sense for the cloud providers to be expanding into this direction. Hyperscalers are already seeing the opportunity to partner with mobile network operators to provide edge services over operators’ low latency 5G networks. AWS launched its product Wavelength, a compute platform dedicated for running on mobile operator’s edge nodes. AWS announced partnerships with Verizon, Vodafone Business, KDDI and SKT. Meanwhile Azure has teamed up with AT&T to provide a similar service that the pair have termed ‘Network Edge Compute’.
However, the hyperscalers’ interest expands beyond partnerships with operators and they have been developing IoT edge computing platforms. AWS, Azure, Google and Ali Cloud all offer platforms that work within their respective cloud environments. These allow enterprises to collect, clean and analyse data in the branch or private data centre, before sending back any relevant data to the public cloud for further storage or analysis. Google has developed Edge TPU and Coral AI chipsets to enable AI workloads to be run in devices or in device gateways, pushing compute even further out.
The next move in edge for the telecoms sector
Mobile operators will face a challenge to grow revenues from the edge computing market. Hyperscalers, IT hardware vendors and independent software vendors have already begun developing their solutions and strategies. Only a few leading operators globally have announced edge initiatives, and even fewer have commercialised services. Further, while AWS, Azure and Google all have well developed communities of developers who are used to working within their environments, if operators choose to go towards proprietary edge compute services, they will need to undergo the daunting task of building a developer community.
Instead operators should work alongside with existing hardware vendors and hyperscale providers to build out the telco edge. While the IT and cloud vendors have the advantage of owning well established computing platforms, the operators have the advantage of 5G low latency networks as well as the benefit of being located near enterprise customers. Further, the edge market will likely end up becoming an extension of the cloud market.
Today, many operators have shied away from directly offering their own clouds. They are moving towards a model where they provide dedicated connectivity to multiple clouds, and tools for organisations to manage all their cloud resources from a single place. Telcos should look to partner with as many edge vendors as possible to build out a similar ecosystem. The next evolution will be from multi-cloud to multi-edge, and operators have a prime opportunity to capitalise now.