Last week, I had the opportunity to attend Edge Computing World where I participated in providing a workshop on the economics of the edge. Over the course of two and a half days of discussions common themes become evident. Here, I share a few observations on the edge computing value proposition and recent market developments that indicate what could be coming in the next few years.
Fuzzy Landscape
To start, I am amazed at market predictions related to the edge opportunity: they range from hundreds of billions to a few trillions of dollars of economic opportunity. Considering the context where industry still needs to define what edge computing is, I wonder the value of such predictions!?
The vast nature of edge computing is what makes it interesting. It brings various parts of the technology ecosystem together: application developers, cloud players, enterprises, device vendors and connectivity providers to name a few. This ecosystem variety manifests itself in the large number of edge computing open source projects leading to a confusing array of overlapping activities.
The Protagonists
There are three main groups in the edge computing landscape: The cloud and data center players, connectivity service providers and end users including enterprises that leverage IoT technologies. To assess the evolution of edge computing, we need to understand the objectives of these players:
- Cloud and data center players
- Extend cloud services to meet the requirements of new applications
- Reduce reliance on connectivity
- Meet regulatory requirements for data localization
- Enterprise players
- Reduce reliance on connectivity to cloud
- Improve application performance
- Meet data localization, security and privacy requirements
- Connectivity players
- Increase network capacity and improve traffic handling by reducing amount of data transport towards core of network
The Solid and Speculative Business Cases
The edge computing value proposition for the cloud and enterprises players is well founded because, at its core, it seeks to reduce reliance on connectivity. The business case could be quantified and benchmarked on that basis to clear the path for the implementation of edge services.
The edge computing value proposition for connectivity players is speculative at the present time. Connectivity players have to distribute their network infrastructure to scale up capacity and reduce the amount of traffic carried to the core network. The business case for edge computing services is contingent on emerging technologies where connectivity players may be bypassed. This increases the risk of investments in telco-related edge services. [There is a case for edge computing in private wireless networks, but the role of the service providers is unclear. See here.]
While connectivity players search for the edge computing killer application, enterprises and cloud players are moving ahead with implementation. Some 5G applications, such as online gaming, could work on today’s networks with adequate performance.
Telco-Cloud Dynamics
Over the past two years, leading telcos have invested in exploring their edge computing value proposition. Virtualization of the telco infrastructure provide telcos the option to offer edge computing services. However, the telco transformation process is iterative and long. Telcos have national scale that limit their geographic reach in comparison with the global cloud players. Recent partnerships between telcos and cloud players – AWS/Verizon and AT&T/Azure – signals resignation of the telcos in relations to edge computing.
In introducing Wavelength, AWS took a step up over Azure and Google Cloud. Wavelength allows telcos – Verizon, SK Telecom, KDDI and Vodafone, with more in the pipeline – to resell AWS services to enterprises. Telcos benefit from keeping traffic local by hosting the same AWS-designed infrastructure. The enterprise benefits from single-digit latency over the same AWS cloud services and operation consistency.
Future Developments
A few areas related to edge computing that we at Xona Partners are looking into for 2020 and the future:
- Low-power compute silicon and memory which promises to be an active area of investment.
- The role of the fixed access service providers who possess fiber infrastructure that well positions them to play in edge computing.
- The orchestration and SW layer for edge computing.
- The efficiency of small data centers proposed to host edge services [see here].
- The application of edge computing in emerging LEO satellite constellations [see here].
- The development of the private networks market [see here].
- The impact of the edge cloud on the major cloud players especially in the context of data localization regulations.