Consumer edge offers growth, but is it being overlooked?

consumer edge computing
Image by Ar_TH | Bigstockphoto

There are many use cases for consumer edge. The majority of consumer digital services are delivered from the cloud, or more accurately, from a combination of cloud and on-device processing. This is true for web, mobile, PC and smart device applications.

Consumer use cases offer a variety of opportunities for telcos and other stakeholders in the area of distributed edge. The latest STL Partners report explores seven such use cases and the implications for telcos.

What is consumer edge?

Generally, edge computing is about moving compute close to where data is generated or consumed, although the actual edge location can vary according to scenarios, use cases and perspectives. It could imply deploying servers per country as opposed to servers per continent. Or it could mean deploying servers in a regional data centre, a telco/ISP network, a stadium, or a home.

Most telco industry focus on edge has been for enterprise use cases (manufacturing, robotics, predictive maintenance, video analytics, etc.) with B2B and B2B2B models. This follows the attention given to private networks and enterprise accounts to find incremental revenues from 5G deployments.

It is notionally easier to envisage migrating enterprise single-site on-premises workloads to telco edge in a single operator-managed environment than to contemplate consumer use cases with inherently more distributed and multi-operator requirements. However, consumer use cases are where we expect to see most of the distributed edge (where processing occurs across multiple edge locations and providers) over the next five years.

Defining the consumer edge

We define “consumer edge” as where edge computing enables or enhances a consumer service or consumer user experience. The edge resources may be located on-premises (e.g. at home or in a public venue), in their connectivity service provider’s network (ISP, fixed, mobile network) or in a regional location (data centre, interconnect or internet exchange) and represent a shift of compute resources either from an end device or out of the traditional cloud.

This is not to say that all the functionality of a use case will “be edge”, as in reality, some or much of the compute will remain on the device and in “the cloud”. Nevertheless, there will be specific rationale and benefit to process some of the experience at the edge (see later).

consumer edge opportunity cloud
Source: STL Partners

Edge nodes are either purchased from an edge compute service provider or implemented by the service provider themselves. The edge resources may be paid for by consumers within a premium service offering (B2C) or paid for by the consumer service provider.

Moving to the consumer edge

There are two broad directions of travel to the edge related to consumer use cases and the underlying compute workloads required to support them – device offload and cloud offload.

Device offload

Consumer digital services are traditionally accessed via smartphones, tablets, PCs, and laptops. Increasingly a whole host of other connected devices need to be considered in the mix; smart TVs, media streaming devices, smart home sensors, wearables (AR glasses, VR headsets, watches), projectors, 360 cameras, etc. Smartphones and tablets alone vary significantly in terms of their compute, storage and display capabilities.

Reaching more device types adds a wider range of connectivity and processing capabilities, memory and electrical power capacity. AR/VR and AI/ML place high demands on processing, which consumes power and generates heat, while adding higher capacity batteries increases form factor and weight. Edge computing can help consumer electronics devices by shifting compute-intensive workloads away from the end device to be processed on a capable edge node and sending the results back to the end device.

Cloud offload

Consumer digital consumer services are traditionally served from the cloud and delivered over the internet. Their servers will be physically located at one or more central data centres around the world operated by the hyperscalers (AWS, MS and Google) or other internet hosting companies. Regular web content is routinely distributed from these central servers closer to consumers using content delivery networks (CDNs). This is fine for static content and reduces latency, server load and data transit, enabling the internet to efficiently deliver vast quantities of content worldwide.

However, for interactive experiences with more complex user interactions requiring server processing, CDNs may not be technically capable and reducing end-end latency becomes even more critical. Additionally, serving the user from far away for high bandwidth interactive experiences can result in inefficient data transit, network congestion and poor user experience. Edge computing can help by offering cloud off-load where certain workloads are moved to an edge node closer to the consumer.

The full report is available here.

Related article: Evolution not revolution: Edge Computing in Hong Kong, China

Be the first to comment

What do you think?

This site uses Akismet to reduce spam. Learn how your comment data is processed.