Telcos versus Hyperscalers in the Race for 5G Edge Computing

5G edge computing provides enterprises and end verticals with the ability to distribute workloads across a wide geographic footprint. This enables local compute processing and single-digit millisecond latency that could not otherwise be obtained with Round-Trip Delays (RTD) associated with traffic to and from public cloud and data center back ends. In addition, the ubiquity and reliability of 5G edge computing serves as the bedrock for continuous operations. In other words, enterprises can operate in a diverse—but not necessarily interconnected—ecosystem while still running production processes and operations in a seamless fashion. Thirdly, 5G edge computing bodes well for local data processing and analytics. This is a fitting value proposition, particularly when we consider that 75% of data in the future will come from edge deployments and terminals. Further, as much as 50% of that data will be processed locally for local decision making and Artificial Intelligence (AI) inference.

5G is the first “G” in cellular technologies that paves the way for Communications Service Providers (CSPs) to operate and work solely in the software layer. For example, Telefonica and Deutsche Telekom are already creating a platform for developers and enterprise verticals to access edge computing resources. Subsequently, 5G edge computing constitutes an opportunity for telecoms to embrace agility, software-driven methodologies, and tool chains (e.g., DevOps, DevSecOps, etc). The genesis of such innovation is the Internet world, and it is new in telecoms. That, along with the potential that 5G edge computing holds for new value creation in the enterprise domain, lays the foundation for new collaboration and co-creation initiatives between niche technology providers, CSPs, and hyperscalers.

5G edge computing promises new growth in the consumer domain but, as the ecosystem matures, new value creation in the enterprise domain abounds. On the technology front, the 5G edge ecosystem is heterogeneous in composition. CSPs have strong connectivity assets like Internet Protocol (IP) transit peering, underlay connectivity (e.g., physical layer, Layer 2.5), and access radio networks (e.g., 5G). But, on the other hand, there is a growing understanding in the industry that CSPs must bolster their capabilities on two fronts: one, modernize their networks with programmable and agile capabilities, and two, better serve the developer community with low-latency edge offerings. CSPs have two options at their disposal: either address those aspects on their own, or invite partners to develop and innovate collaboratively. In regard to the latter, recent partnerships between CSPs and hyperscalers are one market direction that promises to create new demand for 5G edge computing.

The likes of Amazon, Microsoft, and IBM are conscious of these dynamics and have positioned themselves accordingly. With telco-specific solutions like Amazon Web Services (AWS) Wavelength, Microsoft Azure Edge Zones, and IBM Telco Network Cloud, respectively, they can complement CSPs’ connectivity offerings to bring vertical industry solutions to market. These players view themselves as component and service builders that range from scalable cloud infrastructure to high-level platform services with sophisticated data analytics and applications. It is worth noting that hyperscalers do not design for the data center or mass-market. They design for the enterprise customer. For example, Amazon claims that 90% of its roadmap is customer-driven and comes directly from customer feedback.

Enterprises demand on-premises local processing that is secure, preserves data privacy, and can scale in line with enterprises’ global operations. Achieving scale warrants a concerted effort, where CSPs, hyperscalers, technology solution providers (e.g., Enea Openwave, Ericsson, Nokia), and even enterprise verticals work together to bring new partnerships to bear. With the right collaboration arrangements in place, there is no reason why 5G edge computing cannot yield value for all parties involved. But, at present, nobody can predict where this is going to go. It is certainly going to be a captivating journey to see how different parties collaborate and co-create to unlock the value of the 5G edge computing foundations that are being built today.

The industry is at a groundwork stage wherein market players are busy building foundations for 5G edge computing. For example, a great deal of the discourse currently revolves around “new” security models for a distributed environment that extends to core network, transport layers, and edge devices. From a CSP perspective, a question of significance is whether or not they can do more to embrace new ways of working, agility, and software-driven principles. The answer is yes. Are CSPs taking steps in the right direction with 5G edge computing? The answer is a resounding yes. The recent collaboration between Telefonica and Deutsche Telekom is a case in point. Their joint endeavor showcases how CSPs are collaborating by leveraging 5G to build a multi-operator, federated edge cloud. In this case it was a mobile cloud gaming demonstration, a consumer use case. For enterprise, the opportunity is vast, so it may be more optimal for CSPs to go beyond pursuing a singular strategy by collaborating and co-creating with technology providers.

In a collaboration setting, there is a party that leads in terms of business outcomes to be achieved. This notion of co-creation, by contrast, entails two entities sitting around the discussion table as equals to establish, and jointly execute, mutually beneficial goals. IBM’s work with Vodafone is an example of a co-creation initiative. IBM provides vertical integration, and Vodafone provides network capabilities. In general, which party meets Service Level Agreements (SLAs) in a co-creation initiative will depend on context and use case. If the user is consuming a service, then the party responsible for SLAs will be the service provider. Then, in terms of sharing the value that is created, both parties must be innovative to iron out commercial specifics. A good starting point is to start delivering value by having each party play to its strength. CSPs own the connectivity piece, whereas technology providers are agile and innovative. In between, there is an infinite combination of new potential operational models and therein lies the opportunity to collaborate and co-create.

Ultimately, whether it is a collaboration or a co-creation endeavor, CSPs must utilize these industry developments to rebuild themselves while on the road. Firstly, CSPs must be wary of Shiny Object Syndrome (SOS). They should pick the right combination of technologies today that are conducive to easy and quick migration to future technologies tomorrow. Secondly, in ABI Research’s view, CSPs should start to build a cogent case to channel resources that complement their strong network expertise with DevOps and cloud capabilities. This is not trivial, given that it will need to be done in a simultaneous fashion. CSPs will continue to meet the needs of consumers while (re)building their network operations and internal structures to pursue the new enterprise growth that may take some time to unfold.