As the data center industry continues to move IT workloads closer to the end client and further from our core data centers, we are facing complex power, availability, and reliability issues along with new complexities in supply chain, service, skilled workers, and authority having jurisdiction with differing opinions. Peter Panfil, Vice President of Global Power at Vertiv joins us to examine these issues and discuss a new, innovative approach to powering the edge.

Peter leads strategic customer development for the Power business at Vertiv and works to apply the latest power and control technology to proven and emerging topologies to provide the availability, scalability and efficiency levels customers demand. A thirty year data center veteran, he has held executive positions including VP Engineering and VP/GM AC Power prior to his current responsibilities.

Edge: an emerging market

The need for edge infrastructure evolved from the need to improve utilization rates and reduce costs, with larger scale and more intelligent use of redundancy. The recent explosion of content delivery also drove significantly increased demand for low latency and near real-time response rates. Customers are now expecting this real-time response in many, if not all of their IT applications.

While the edge is an effective solution, it does have some practical limitations, things like the speed of light. We also know more machine to machine applications and systems, like autonomous vehicles, will be almost entirely reliant on the edge. The types of applications, physics, economics and ‘law of the land’ all have impacts on the edge. Panfil stresses that the edge is not just a scaled down core; the environments are fundamentally different.

The core is typically designed with personnel access in mind, at the edge, this is not always possible. Panfil says the edge needs to have seamless operation of equipment to remain a reliable, and viable option for customers. As such, the edge is also expected to be delivered already integrated and performance validated.

The evolving edge environment

Panfil says that the edge should be an all in one solution that can be dropped into place. This is obviously a fundamentally different approach to how we look at, and produce core infrastructure. This requires innovation from top to bottom - there needs to be champions in every part of the product. Additionally, safety will need to be at the forefront - ensuring equipment meets the highest industry standards.

To achieve all this, third parties will need to be brought into the process of production earlier, making the process easier. Moving from a project approach to a product approach will improve efficiency in the deployment of edge. The more normalized deployments get, the less validation is required, resulting in reduced variability and minimized needs for oversight. All of this results in increased speed - a key demand of many customers across the industry.

Solutions need talent

Panfil says that 30% of data center professionals plan to have left the industry by 2025, a large portion of these leaving due to retirement. This labor shortage will have significant impacts on markets in North America, but will also impact the industry as a whole. System design and deployment experience becomes a premium skill in this environment. Yet, this also opens up opportunities for remote monitoring and data driven maintenance, which will become more critical over time.

What matters at the edge?

Reliability

If the edge goes down, the core goes down too. Edge sites need to be reliable to ensure continuous services for customers across the board. Edge sites will often be hard, or impossible to access due to their location, so the equipment will need to be as close as possible to 100% autonomy, Additionally, communication about potential downtime or interruptions to services need to be communicated before they occur.

Connectivity

Remote diagnostics become absolutely essential, as enhanced connectivity results in monitoring by experts who can prevent, diagnose and solve problems.

Scalability

Many core data centers are bespoke, designed unique to each customer and/or project. Scale in the edge realm will allow normalization and innovation. The more scalable, the lower costs may be overall. Regional variants can be taken into account in the normalization process.

Expertise

At its core, vision and expertise contribute to innovation. Working on ways to improve infrastructure and technology will prove helpful to edge rollout. Finally, understanding limits helps understanding how to optimize performance.

Make sure you tune into the free, on-demand session with Peter Panfil to hear his predictions for the future, and the Q&A segment from the live recording.