As featured in Acceleration Economy Network

It is now commonplace to hear folks say that edge computing does not exist or the edge doesn’t matter. It is meaningless and we shouldn’t talk about it. To be fair, for most audiences, edge computing and the idea of the “edge” does not matter. It is complicated, confusing stuff that most folks just won’t grasp, much like cloud. Much like AI. Most folks don’t care as long as the application works.

But much like cloud, AI, and other hyped and perennially misunderstood and confused techno trends, edge computing matters. It’s not that we shouldn’t talk about any of these things. We should talk about them in a more consistent and informed manner.

Why is edge infrastructure important?

Edge infrastructure is a vital part of the edge computing equation. Why? Because edge computing is about distributed computing architectures and technologies that will enable new types of applications that were not possible before. It is happening and evolving due to the continued advancement of digital technologies such as Edge AI, and cloud-native architectures and technologies. 

What do we exactly mean by edge infrastructure? It’s the deployment of compute define as processing, memory, data storage, and network resources in proximity of an edge environment, which can be a campus, a factory, a home, or a person.

Edge infrastructure provides shareable compute resources that can be leveraged by an edge application. This typically means that an application and data can be distributed between the edge infrastructure and endpoint devices to enable new application architectures through resource and performance augmentation. Think in terms of the X-tier architecture mentioned in the “How the Edge Cloud is Shaping the Future of Computing” installment of the Cutting Edge series on Acceleration Economy.

So, no, edge infrastructure isn’t a new pivot because edge computing hasn’t taken off or because we have been doing edge computing forever. It is becoming the home of edge clouds that are proliferating across the edge. There is real, new stuff going on here that we should talk about. 

How is edge infrastructure changing?

The single most influential thing that is changing the way we think of edge infrastructure is cloud-native technologies. We have covered these technologies before but here is a reminder:

• Virtualization – is driving the software-defining of hardware and physical infrastructure.
• Containerization – is granulizing the way we can package and consume compute resources.
• Microservice-based Architectures & Software Designs – are changing the way we design software in a more “service-oriented” way. Think of it as Service Oriented Architecture having a baby with Web Services.

As these cloud principles make their way toward the “edge”, they are driving a convergence of communications technologies and data center technologies and operational tools that are coming together at the edge under the umbrella of ”cloud-native” spawning all the varieties of edge clouds that we see emerging today such as:

• Telco Cloud
• Multi-Access Edge Cloud
• Private Enterprise Edge Cloud
• Industrial Edge Cloud
• And more to come

The edge demands new capabilities across locationally distributed compute resources that we don’t see in centralized clusters of compute resources in hyperscale data centers. Some of these new capabilities that are unique to edge cloud infrastructure are:

• Edge Orchestration 
• Edge Service Management
• Edge Analytics & Optimization 

It is true, these are not unique concepts to either cloud or the telco camps. It is also true that these concepts have not been figured out for new edge infrastructure scenarios. We are seeing a hybrid of technologies and methods that are developing as vendors across the board are looking help their customers figure out the net new in edge infrastructure.

Once again, Open RAN presents a great example of edge computing innovation. One particular element of O-RAN is the RIC or RAN intelligent controller which functions in the Open RAN architecture to manage and optimize RAN resources and enable the intelligent delivery of mobile network services across an operator’s network infrastructure. The exciting things about the RIC are the rApps for non-real time network functions that can execute with more than a second in latency, and xApps for real-time network functions that can be built on top of a RIC to infuse intelligence and autonomous features into the edge infrastructure and its operation.

Key Takeaways for the C-Suite

Edge infrastructure and what promises to be its dramatic evolution over the coming years is not a myth. It isn’t hype. It is happening today with many technologies emerging to drive this fascinating reinvention of digital infrastructure.  

This means that technology leaders in your firm shouldn’t use the excuse that IT folks don’t know CT so they should just depend on the folks who know telecom technologies. Sure, you can assume that will be a sustainable mindset if these two world stay separate without consequence. That is not the case. 

Your CIO, CTO, CISO, your Chief Digital X Officer, should know this stuff or they will lose relevance in your organization over time. If innovation and competitiveness truly matter to your organization, you will want technology leaders at the helm who know about this convergence of technologies and infrastructures that will be the source of much of the innovation shaping the future of cloud and the edge.