What Traditional Data Center Service Providers Should Consider When Developing an Edge Strategy

Mike Michalik, CTO for Evoque Data Centers Solutions

A common question we hear these days from potential colocation customers is, “What is your strategy for providing solutions for edge computing?” This tells us that today, our customers are looking to traditional data center providers to help them solve the problems of edge deployment.

As in the early days of cloud, enterprises are now in a “wait-and-see” mode regarding edge. While enterprises have an immediate need to solve for data capture, compute, low latency, greater cost efficiencies, and optimized end user experiences, they are waiting to see how edge ecosystems will develop and how edge technologies will evolve. Over the next few years, we will continue to see different “flavors” of edge technology emerge, just as different “flavors” of cloud have emerged.

At this early stage, many colocation providers are defining their position in the developing edge marketplace. There will always be a need for traditional facilities to house centralized IT footprints at the core. But data center providers need to develop an edge strategy – a plan for how they will solve for the push to highly distributed architectures and provide services for customers who are seeking edge solutions, wherever that edge may be.

Drew Leonard, VP Strategy,
Evoque Data Center Solutions

Understanding Edge Needs for Customers and Industries

Building an edge strategy requires an understanding of edge needs from a market perspective. The basic premise of edge computing is the need to push compute and content out to the network edge. The dilemma for data center providers is solving for the multitude of use cases from enterprises. Organizations in different industries will deploy different workloads, requiring different types of edge ecosystems. The individual needs of each industry and company will determine the edge solutions and configurations they require.

The primary purpose of edge data centers is to reduce latency by bringing servers and networks closer to end users or data sources. But different companies will have different latency requirements. Autonomous vehicle providers, for example, require real-time application delivery to ensure the safety of vehicle passengers, other motorists, and pedestrians. Agricultural conglomerates, on the other hand, may only use edge data centers for IoT functions such as regional data collection and low latency data processing.

Flexibility will be the key to success in delivering solutions that meet customers’ needs and requirements. No one solution will accommodate all users, but being in a position to react to evolving market needs will go a long way.

Today, most edge data center pioneers are adapting current data center designs into smaller packages. This includes hardened facilities with robust 5-9’s uptime SLAs or better, which can be extremely costly. To scale effectively and meet the needs for a hyper-distributed architecture, traditional data center providers will need to innovate and be creative to deliver more cost-effective power solutions with acceptable levels of resiliency. As enterprises evolve their edge strategies, they will quickly find a balance between more economical edge deployments in lieu of higher risk of downtime.

Flexibility will be the key to success in delivering solutions that meet customers’ needs and requirements. No one solution will accommodate all users, but being in a position to react to evolving market needs will go a long way.

Developing Vertical-Based Solutions

Data center providers will need to develop edge strategies based on the requirements of the industry verticals and niches they serve. For example, content providers will require edge ecosystems with many points of presence to deliver high-end user experiences to consumer devices (homes, phones, tablets, etc.) using content-rich workloads. Autonomous cars, because of the potential safety risks, will require many data gathering and processing locations to ensure almost zero latency in both dense population centers and rural areas. The healthcare industry will require global sensors and data gathering locations to capture the constant flow of information transmitted by personal IoT medical devices.

Each of these illustrative verticals will also have edge needs that are not as latency sensitive, such as collecting and processing user profile data, technology performance, ordering and billing of services, etc. But hyper-distributed edge architecture will still be required to service each vertical.

The basic principles of the edge data center are based on several key factors: location, interconnectivity, smart infrastructures, and cost. Basic use cases are based on mission criticality, cost effectiveness, revenue generation, scalability, and end user experience.

The edge strategy requires an appropriate investment in the right facilities in the right locations. It’s been said that “edge is where the customer needs it to be,” so providers should be willing to go where the customer wants to go.  Just as important is how the provider will deliver the right technologies to ensure appropriate levels of connectivity, uptime, reliability, and security, both at central facilities and edge locations. Providers must develop the necessary partnerships to deliver the edge technologies that they can’t build in-house.  Partner companies may include physical infrastructure providers (containerized or micro data centers), traditional, wireless, and software-defined network (SDN) providers, antenna and tower companies, local utilities, and technology providers for specialized industries (i.e. developers of IoT application software for use in edge ecosystems).

Regardless of where you sit as a provider or end user, the edge represents an opportunity for growth. Data center providers that offer scalable solutions to help their customers solve their edge deployment problems will see pull-through business into their core facilities from these customers. Data center users always find value in working with providers that can meet all of their needs.  As always, the more value a provider can offer in edge solutions, the more customers will come their way.

Drew Leonard is VP of Strategy for Evoque Data Centers Solutions. Mike Michalik is CTO for Evoque Data Centers Solutions.