With the emergence of 5G and the internet of things, the enterprise edge is being developed to accommodate massive scale. After all, when virtually everything we see and touch is connected to a global digital ecosystem, the one thing you can’t skimp on is scale.

But as we learned with the cloud, great scale begets great costs, even if the price per transmitted bit or stored byte is dramatically lower than previous generations of infrastructure. For organizations looking to stay on the forefront of this emerging data frontier, the overarching focus should be to enable a fully scalable edge within the lowest possible cost envelope.

Expanding footprint

According to Grand View Research, spending on the global edge is expected to top $61 billion by 2023, growing 38.4% per year on average. A key factor in this growth is the level of artificial intelligence that is expected to permeate the infrastructure, which in turn should help control costs by enabling real-time decision-making, dynamic resource allocation and other automated processes that should make the edge more responsive and easier to manage. In large part, this will facilitate high-speed performance and tight ratios between bandwidth and data loads.

To take advantage of these tools, however, network operators must be able to deploy high-bandwidth, low-power, small-form-factor hardware solutions in high volumes and at manageable price points. Demarcation devices, for one, should operate efficiently at 1 and 10Gbit/s rates, while at the same time provide zero-touch provisioning to enable rapid expansion of network services. An added bonus is a rugged design that can withstand temperature extremes without additional cooling or other support infrastructure.

Another key element in edge networking is timing. Many 5G fronthaul applications require nanosecond-level synchronization, but most grandmaster clock solutions today are external to antenna/receiver assemblies, increasing both the cost and space requirements at network access sites. Again, small form factor is crucial here, as is the ability to provide both electrical and optical interfaces to replace traditional RF feeds that are both costly and complex.

Under an integrated virtual network, organizations can manage their diverse collections of compute and storage as a single entity.

But perhaps the most effective means to ensure edge costs don’t spiral out of the control is effective workload management. The edge is not alone in the connected digital universe; it is backed up by both the cloud and the data center. Utilizing all facets of data infrastructure can help manage costs across the board, but it has to be done as seamlessly as possible and with the highest degree of flexibility to ensure new services can be deployed quickly to meet changing demand.

The most effective way to do this is by deploying cutting edge management and orchestration on the network virtualization software layer. Under an integrated virtual network, organizations can manage their diverse collections of compute and storage as a single entity, ensuring that workloads are balanced across the entire infrastructure so as not to force unnecessary hardware provisioning or expansion on any one facet of the environment. At the same time, it allows for highly precise data handling so that the most urgent applications, such as autonomous vehicles or health system monitoring, are pushed as far out on the edge as possible to enable high-speed service, while less critical data can be collected for longer term processing and analytics.

The edge is already a crucial component of the expanding enterprise digital footprint, and it would be a shame if it were to fall to the same careless provisioning and lack of planning that hampers most data center these days. But by adopting efficiency and flexibility as core elements right from the start, most organizations should find themselves well positioned to take full advantage of 5G and the IoT – and whatever comes after.