Edge native is a challenging concept to explain because there are a growing number of vendor descriptions about what edge-native is. It’s typically used to tout the nuances and differentiations of a particular vendors offering. How will it matter and how will it impact the future of computing at the edge? 

What is edge native?

To understand edge native, we need to understand cloud native. Cloud native a term used to describe technologies and architectures used in cloud computing, namely, microservices architecture, containerization/virtualization, and now serverless. These aspects collectively enable the paradigm of software development, infrastructure management, and the Everything-as-a-Service IT delivery models that we have all grown to love over the last 15 years with cloud.

Edge-native is a twist on cloud-native but factors in the very important nuances and challenges that come with extending cloud computing and cloud app development toward the edge. In simple terms, it involves a little of the same cloud computing technologies and architectures with some new permutations to deal with the particulars of edge deployments of compute infrastructure, delivery of IT services, and software development.

The whole idea of edge-native is to expose the functions of a software-defined network to developers who can build network and edge resource-aware applications that can be placed optimally across an edge infrastructure.

The idea of placement is important because the edge-native application workload and data need to be deployed at an edge node location that can feasibly and optimally deliver the application services with the quality of service required of the application. This is particularly important for applications that require low latency but more importantly high reliability.

Here’s how you should look at edge native in simple terms.

• From an infrastructure perspective, it means dynamic management, orchestration, and placement of application images, workloads, and data across an edge infrastructure, which could be a geographically dispersed network of edge clouds or cloudlets.
• From an application perspective – edge native enables a new breed of distributed applications that can take advantage of edge cloud infrastructures with network, location, and compute resource awareness.

There are challenges with building edge-native applications:

• Resource diversity is not the same as the cloud. One of the things you have probably noticed is the fast-growing range of instances and services offered by your favorite public cloud provider. It’s not that easy or economical to host such a diversity of infrastructure, hardware, software, and instance options the further toward the edge you go. The economics don’t scale the same as it might in the megaplex data centers of cloud service providers, although those economics may be changing.
• Portability is an issue. Heterogenous edge clouds and traditional telco and cloud silos present obstacles for applications to transverse the edge. Much of this is due to a bevy of interoperability nuances and issues both software and hardware that could make an application an unsupportable tenant on a cloudlet that might just happen to be the only one close enough to the client endpoint device to deliver on the requisite quality of service on demand or QoD. It could be the case that the required processor or accelerator is not available. 

What does edge native mean for the future of edge computing?

Edge native will be relevant mostly for edge infrastructures supporting mobile applications on a public mobile network or private industrial environment. There could be opportunities in enterprise but less likely at the moment.

For ideal candidate applications and infrastructures, edge native will:

• Firstly, change the way we manage workloads across a cluster of edge clouds. This means new tools, architectures, operational practices, and investment strategies. As I mentioned earlier, the edge node hosted in a co-lo site or in a baseband cabinet under a cell tower is not your CIO’s or cloud provider’s megaplex data center.
• Second, it will change the way applications will be developed on edge infrastructure. It will technologically be cloud-native, and deployed in a fashion that is location and context aware and is mindful of infrastructure service quality and assurance. Why? It’s likely that most of these edge-native applications are going to rely on highly reliable, deterministic connectivity between the edge infrastructure and the endpoint devices it supports. Yes, latency will matter, but you can’t get that without reliability. 

Is this futuristic technology? No. Much of the tools and platforms are being developed by familiar players such as VMware, IBM Hat, Wind River, Ericsson, Nokia, hyperscalers such as Microsoft, AWS, Meta, and Google, and a ton of smaller players such as ZEDEDA.   

What does this mean for the C-Suite?

Developers and systems architects and CTOs should get really excited. CIOs might be a little concerned as edge native is another game when they just started to get used to cloud computing. 

The bottom line is that there will be a new learning curve. Conventional notions of cloud computing don’t translate perfectly across the edges. There are edge native nuances that will eventually need to be understood to unleash the potential and the promise of new edge computing paradigms. 

Does that mean the cloud is dead. Absolutely not. Even in most of these new edge applications, central clouds can play an idea role in providing central management of edge resources, networks, and infrastructures, as well as a central registry and repository for applications and data that will be placed across edge cloud infrastructures.