11 July 2022

By Simon Michie, CTO, Pulsant

Edge computing’s radical departure from the standard cloud model presents service-providers and application builders with new markets and a platform for innovation.

Bringing computing power to the edge of the cloud, close to where organisations generate and use it, accelerates digital transformation, regardless of an organisation’s location. In the UK, the edge gives businesses beyond the South East, where the hyperscalers are situated, access to low latency digital services, enabling them to use a vast arsenal of powerful new SaaS applications.

For SaaS vendors, the edge opens the door to millions more customers and latest subscription and delivery models. Application-builders and providers can orchestrate the network functions and computational infrastructure needed for secure delivery to their end consumers. Content providers are able to offload from their central servers, transforming access and efficiency.

We already see exciting use cases supported by 5G-enabled edge computing. Including remote monitoring and diagnostics in healthcare along with 3D medical imaging, while the availability of low-cost, easily-deployed sensors extends the internet of things (IoT) and Industry 4.0 applications that include artificial intelligence (AI) and flexible automation requiring near real-time responses. In manufacturing, automated, high complexity production will be possible in new locations. In port management and logistics, edge computing delivers major gains in efficiency across expansive campuses and along supply chains. Multi-party gaming, advanced drone use, AR and VR applications are all enabled by edge computing.

The potential is such that Statista estimates the worldwide edge market will grow to $250.6bn as soon as 2024. By 2030, the number of devices connected to the internet could reach 125 billion according to IHS Markit.

This dramatic growth depends on low latency. Augmented reality, virtual reality and most Industry 4.0 IoT implementations require latency as low as 1ms to 10ms. Edge-located machine learning (ML) models also require low latency once they are trained on masses of data in the cloud. As well as enabling organisations to deploy ML when they otherwise could not, edge computing reduces the costs of backhaul to the main hyperscalers’ hubs.

However, the success of edge computing requires service-providers and end-users to have access to a scalable national network of edge data centres to ensure that, regardless of location, they have low latency, high-bandwidth connections. The more advanced edge platforms achieve sub 5ms latency, using a networking approach that spreads the load across several regional data centres. They not only process data closer to each end-user but reduce the congestion in backhaul that undermines performance.

Edge data centre networks must have high-speed connectivity to cloud services so transfers of data to the central hub are seamless, with route diversity to ensure resilience in the event of outages. And as more enterprises use edge-reliant applications, there must be sufficient capacity to meet future demand for bandwidth.

Partnering with one provider with a nationwide network of cloud and data centres, rather than a collection of multiple providers will also deliver benefits in terms of security and remove the complexities of integrating IT services. If solution-builders and service-providers are serious about expanding into all corners of the country their customers will also need access to a platform that has a good geographical spread of strategically-located sites that are purpose-built, and which maximise coverage in any area.

Since many organisations will balance their workloads between the public, private clouds and the edge for reasons of flexibility, security and cost, edge data centres need to facilitate hybrid patterns of use and multi-cloud models. They should sit between the on-ramp to the public cloud and network-to-network interfaces with the telecoms providers. This functioning edge ecosystem also needs to include specialists in microservices, containerisation, virtualisation and related fields.

We can see how ecosystems are developing, as telecoms and hyperscalers form partnerships with existing edge computing platforms. Last year, for example, Telefonica Germany partnered with AWS and Ericsson to virtualise its 5G core network. The purpose is to enable quick integration of new applications, removing the need for time-consuming hardware set-ups while reducing costs.

The edge also requires partnerships between the big-name public cloud vendors and edge providers so that organisations deploying advanced applications have the flexibility to locate workloads where they work best. Service-providers should be looking out for edge providers that have these relationships in place as well as the vital attribute of ubiquitous low latency coverage and high-speed fibre connections between centres.

For service-providers and solution-builders, the new maturity of edge computing opens the main gateway to new markets and the ability to create new products and business models that deliver higher levels of service and increased revenues. Yet to be successful, it still requires a smart selection of partners. The right edge computing platform will transform the fortunes of SaaS providers and put them at the centre of digital transformation for thousands of new customers.

The bottom line is that through edge data centres, businesses can easily power their growth, their entry into new markets and delivery of new services to almost anywhere, therefore meaning solution-builders can provide products and services built on analytics, AI and machine learning.