Welcome to the Cutting Edge. I’m Leonard Lee, the Acceleration Economy Network analyst who will be taking you on a journey to the edge of the network, the IoT edge to explore and discover what all the excitement is about edge computing. A little about my background – I’ve been in technology and consulting for almost 30 years. I was formerly a managing partner at Gartner Inc, and partner and principal at a number of management and technology consulting firms including IBM, PwC and EY. Through my firm, neXt Curve, I have been working with some of the leading companies across sectors of the technology industry on the emerging topic of edge computing.
What is edge computing? The term has become the buzzword du jour. Many industry pundits claim that edge computing is the next big thing. Others claim that we have been doing computing at the edge for decades. To make matters even more confusing, some folks say that the edge is everything outside of the data center. What does it all mean?
In this series of analyst contributions to the Acceleration Economy, I want to help you arrive at a clear understanding of what edge computing is, the different forms it takes on, what it means for your business, and why you should care. We will also explore what companies and industries are doing to extract competitive advantage by innovating on top of evolving edge computing technologies, applications, and business models.
Edge Computing: Where the Action is
For a few years now, I’ve been looking at the topic of edge computing with interest from the emergence of CDN’s (Content Distribution Networks) back in the first Dotcom era. But in recent years, I see it as a convergence point where IT, telecommunications (CT), and various operational domains (OT) are coming together. Think of it as the cloud meets 5G, meets IoT in all its different flavors. Of course, we can’t forget emerging AI computing and hybrid software.
We are also seeing a wide range of technology trends and themes intersecting at all points across the “edge”. Old methods are met with new ways to think about compute (how we deploy ICT resources)and computing (how we architect and deploy software applications and data). The space is being shaped by leading-edge data center technologies and endpoint technologies from software, to hardware, down to the silicon.
The best part of edge computing is that everyone is simultaneously exploring its capabilities and potential. It’s a topic ripe with both business value and technical challenges for technologists and business leaders.
Words Have Meaning, So Does The Edge
I want to begin this column by addressing semantics because edge computing suffers from rampant misconceptions and misrepresentations. I can’t count the number of analyst forums on edge computing where everyone was talking past each other. Establishing grounded nomenclature is an important first step in any hyped technology discourse because words have meaning. We can’t get into a fruitful discussion about how it can make a difference for your business without being clear on what we are saying to each other.
So, let’s start with the edge. What is it? In the past five years, I have had endless debates on this word with technology leaders from IT, CT and OT. There are domain-specific nuances to what the edge is for each of these communities, but holistically speaking the edge is the default location. It is simply where a “client” endpoint device is located. That location can be the network edge, a factory floor, your home or you.
It’s then important to understand that edge computing is about distributed computing. It’s not standalone computing on your smartphone or your smart fridge. In my opinion, this is an important distinction because there are many technologies and innovations making on-device computing more compelling alternative to distributed computing. In the context of edge computing, the endpoint device is a client. In the case of a smartphone, it happens to be an increasingly powerful one. However, an endpoint device is an essential participant in an edge computing application or system design much like the smartphone is a major client in cloud computing of today.
For technically savvy folks, edge computing is essentially a client-server computing model. The edge compute (the resource for processing, storage, and networking) hosts the server that is in the proximity of a client device located at a default location. The server can be located on an intelligent IoT gateway, an edge cloud node at a local cell tower, or a branch office of a telco operator.
What differentiates edge computing from “cloud computing” is that the server is located anywhere but a hyperscale public cloud or enterprise data center. The edge server is placed closer to the default location for various reasons like latency or data sovereignty. Furthermore, the technology advancements are creating new possibilities for designing distributed systems and applications at the “edge” that present compelling alternatives to centralized cloud computing.
Since the advent of CDNs, edge computing has been a counterpoint to the centralized cloud computing that has been the vogue in the IT world for the past decade. Technologically, the recent excitement about edge computing is, ironically, largely attributable to advancements in cloud computing itself over the past 7 years with the advent of cloud-native architectures. The hyperscalers such as AWS, Google, and Microsoft have taken notice. We will get into that mind-bending topic in a separate article as it deserves dedicated attention.
Framing The Edge
Edge and edge computing are extremely diverse topics. Certainly more diverse than the conventional notion of cloud computing. Why? It is more than the principle of densifying as much compute as possible into hyperscale data centers that host massive public clouds. Essentially, the edge is a wide variety of locations and edge computing is a fast-growing range of architectures and modalities that cater to localized deployment requirements.
Edge computing comes with a broad and deep range of complexities that are generally misunderstood by some or insufficiently appreciated by others. It is not a domain that is well served by abstractions. The details and complexities matter. You will have to address and navigate them in order to discover and realize value for your business whether you are a service provider or an enterprise end-user.
The diagram below is one that I put together to frame the different layers and viewpoints of edge and edge computing.
I will be using this taxonomy and blueprint as a reference in our ongoing discussions. Together, we can refine it for those who venture to the cutting edge of our digital future.
There is a tremendous amount to unpack and digest. It may take time to understand the nomenclature, the taxonomy and concepts that underpin edge computing. Honestly, it also took me a while to understand too, but this column will get you there. We will get there together.