And you thought the cloud was complicated. When cloud computing first debuted, everyone was trying to figure out what it was; it meant different things to different people. Fast-forward to 2016, and it was déjà vu all over again – this time with the edge. In fact, one can argue that edge computing is even more ambiguous, misunderstood, and unknown at this point than the cloud was when it started out. Case in point: one customer recently told me that edge “means everything and nothing.”
A big part of that uncertainty stems from the fact that the edge is a simple word for a completely emerging, complex, and misunderstood topic. It leaves people asking what’s real in this space, what are the basic requirements, how do others define their edge strategies, etc. But with Forbes Magazine naming edge computing one of the Top 10 trends for digital transformation in 2018, it’s a topic that companies who want to stay competitive cannot afford to ignore.
While the industry has not yet landed on a completely cohesive definition of the edge, there has been some basic alignment. In general, it’s about moving compute closer to where data is being generated — where data and content is valuable, mobile and distributed. Often times this data requires information processing for real-time decision-making. Rather than incur the cost and latency of sending this information to the cloud or centralized data center, many businesses are looking to incorporate edge computing within their infrastructures.
Take for example the airline industry. AviationWeek reported some airplanes have 5,000 sensors that generate up to 10GB of data per second. A single twin-engine aircraft with a 12-hour flight can produce up to 844TB of data. Even with a 100GB pipe, it would take over 18 hours to transfer this data to a centralized data center. With edge computing, you can speed analysis of the critical data and ensure the plane is ready to fly again, while back-hauling the rest of the data.
ESI Takes on Edge Computing
If you didn’t attend Mobile World Congress Barcelona or Dell Technologies World, you probably missed our Extreme Scale Infrastructure (ESI) team talking about the latest developments with edge computing. Here are some of the key themes we’ve been sharing at events:
Applying Hyperscale Principles to the Edge: It may sound counter-intuitive, but as Dell vice president and fellow Ty Schmitt explains in this interview at Dell Technologies World 2018, ESI is taking what we’ve learned building Modular Data Centers (MDCs) for macro-data centers and is now applying those principles to edge deployments. It’s still about packaging IT, power, cooling, security and management into a right-sized solution for the customer, but it’s the operational model and approach that has changed. For example, with edge, you may be trying to figure out how to distribute 10 megawatts of power for distributed use across a thousand different locations as opposed to concentrating it in a single, central location. Watch to learn more.
ESI’s Latest Micro MDC Concept Takes Center Stage: Companies are realizing that trying to migrate all of their data to the cloud or centralized data center generates bandwidth congestion, latency and cost. In many cases it’s just not fast enough (see above airline example or think self-driving cars and the billions of other smart devices woven into the IoT landscape). That’s why ESI has introduced the idea of a micro MDC, designed specifically for the edge.
Previewing its latest prototype at Mobile World Congress 2018, Dell’s distinguished engineer Mark Bailey explains how the tenets of IT, power, cooling and a structure inclusive of security, networking and power distribution were incorporated into a single, 18 kilowatt, one-rack solution. Dive deeper into the specifics with Mark here.
Managing at the Edge: So now that we’ve talked about moving data centers out to the edge, management can’t be an after-thought. With potentially thousands of micro MDCs spread across the world – often times in remote locations and without operators onsite – how do you manage? Dell’s senior principal engineer Tyler Duncan describes how we’re monitoring and managing at the edge at the recent OSIsoft PI World conference in San Francisco. Watch the video to learn how we’re integrating OSIsoft’s PI system into our micro MDCs to monitor electricity consumption, avoid peak power charges, detect early warning signs of failures and other tasks.
Conclusion
With the space still emerging, companies that want to occupy the edge are going to face logistical, security, regulatory challenges and more. Organizations are going to need a partner or provider who will take the time to learn their strategy and take a cross-platform view, with specialists from facility, IT, logistical, network and security arenas.
While ESI has been busy talking to customers about edge computing from an IT and data center perspective, it’s just one piece of the puzzle, with software and services equally as important. Dell Technologies and Dell knows that edge solutions are evolving and we are investing tremendous resources throughout the entire company to create products and solutions, proof of concepts and field trials to ensure we meet customer your requirements at the edge.
Visit us here if you’d like to learn more about ESI capabilities or reach out to ESI@dell.com if you want to discuss your edge initiatives.
