How Edge Computing Bridges the Past and Future
“Edge computing is nothing new.”
“Edge computing is a whole new way of conducting business.”
Opinions swing from one end of the spectrum to another. The truth is somewhere in between. Like “something old, something new, something borrowed, and something blue,” edge computing borrows from the past but is propelling us toward a safer, easier, and more sustainable future.
To understand how edge computing bridges the past to the future, let’s start with what we mean by “edge computing.” Edge computing consists of distributed systems made up of compute, storage, and networking running applications that enable enterprises to act on data near its point of creation. At Dell Technologies, we often talk about two types of edges, near and far. These definitions describe logically how close to the data that applications can run. They also frame infrastructure constraints such as space, power and cooling, dust, vibration, shock, temperature extremes, connectivity, security, and so on, that enterprises must consider at their edge.
Let’s look at these types of locations and use cases.
Near edge – These deployments typically deliver a turnaround time under 75 milliseconds with compute and storage located between the far edge and the central data center. An example of a near edge deployment is video surveillance, in which video is analyzed in real-time instead of being sent to a central location and wasting precious time. Real-time data can help identify time-sensitive events such as potential thefts and more. With this type of solution, enterprises can act quickly and increase the value from the data collected by cameras and other sensors. At the near edge, enterprises have some space and freedom to expand the compute power and application footprint vs. in far edge deployments.
Far edge – These deployments typically give enterprises turnaround time less than 20 milliseconds. Compute is located as closely as possible to the point of data generation (endpoint). For example, in manufacturing, enterprises deploy a compute node at the far edge to collect, aggregate, and translate data from machinery on the factory floor. That compute node then sends the data for processing in a near edge or a central location. Far edge deployments are characterized by many constraints that can be readily overcome with the correct software, infrastructure, and connectivity.
Now that we can define it, we can answer the original question about how edge bridges from the past to the future:
Something old – A colleague of mine says that “Edge is the oldest emerging technology” because it keeps changing and challenging us in new ways. If we go back in time, in 1984, IBM released the 5531 Industrial Computer. Some can argue that it was an edge solution because enterprises used it to process data near the point of data creation. It differs from today’s edge computing because the 5531 was a monolithic system that solved a single problem for a single site or location. Edge computing today is more about solving distributed application challenges than providing a single-use case for a single place.
The term “edge” was coined back in the 1990s when Akamai created the first content delivery network (CDN). Since then, we’ve seen a rapid evolution of edge beyond content delivery.
Something new – Data types and volumes, application requirements, security, and harsh environments are new aspects of edge deployments. As customers push the limits to get more compute power closer to endpoints and systems become smaller and more robust, customers will deploy more compute in their edge sites. They will require systems to operate at higher performance levels in locations that do not meet the cooling specifications of a data center. Examples are agricultural or oil and gas fields and factory floors with significantly higher levels of operating temperatures and vibrations than found in a data center.
The constant push of IT limits and increased adoption means we need to evolve our solutions to overcome the new edge constraints. Some of those constraints are in data management, security, and serviceability challenges.
Something borrowed – Telecommunications service providers lived at the edge even before “edge computing” was a common phrase. Suppose we define a mobile device as an endpoint for data collection. In that case, the cell tower is the closest point to data collection. A cell site is the edge of telecom because it is the nearest point to the endpoints. Although the concept existed in the telecommunications environment, the improvements in technology and the demand from enterprises created new edge types even for telecommunications service providers. One example is the emergence of Multi-Access Edge Computing (MEC) solutions that enable application developers and content providers to use cloud capabilities at the edge of a network.
Something blue – The color blue symbolizes many things. Two of the most common are “imagination” and “trust.” With the rise of AI and machine learning and the need to gain faster insights from the data, enterprises across every industry are re-imagining and implementing new use cases that impact the triple-bottom-line – social, environmental (or ecological), and financial. And they’re doing so with solutions built on best-of-breed software and hardware with intrinsic security built into every layer to defend against internal and external threats.
The fact is edge computing is a combination of all these things and requires learning from the past to be ready for a future. It is an emerging deployment with enormous potential that requires expertise and optimized infrastructure to generate insights where and when needed, with built-in security for an expanding data and infrastructure landscape. And because this evolving landscape can lead to sprawl, it demands careful planning to consolidate operations, data and infrastructure to achieve target outcomes. We are here to help our customers and partners simplify their edge and create a bridge to a safer, easier, and more sustainable future.
To learn more about edge computing, visit our edge website.