David Linthicum is Unlocking the Power of Edge Computing
Chief cloud strategy officer, David Linthicum, lives life on the edge.
The edge of the network, that is.
Since the early days of cloud computing, information on the network has been processed in one, centralized location. But Linthicum has spent the last few years understanding how we can move that data and computing power closer to where the emerging slew of connected applications need it: the edge.
Here’s an oft-cited example to support why Linthicum is so edge focused: self-driving vehicles, a technological advancement that is moving closer to widespread adoption, can’t wait for on-board computers to query the cloud when deciding how to respond to an oncoming vehicle 50 feet away. The round-trip for the data, plus the extra in-cloud processing time, would simply take too long to get an answer back—not to mention, the wireless connection might fail.
By taking data processing to the edge of the network and moving it closer to the source of the data, the car can make a decision much more quickly. The edge is, quite literally, providing a life or death solution.
Linthicum predicts that edge computing will have a wide range of applications beyond this single example, changing the way we build tomorrow’s enterprise computing systems. We asked him to explain how edge computing could continue to usher in a new era of efficiency, reliability and speed—and make emergent technologies like autonomous vehicles, smart drones and autonomous remote industrial plants increasingly common.
Edge Q&A with David Linthicum
(Responses have been lightly edited for clarity).
Can you explain why edge computing is so important?
Edge extends our ability to put data collection closer to the source. That’s an architectural advantage because we don’t have to deal with as much latency. We can make decisions closer to the data instantaneously, without having to communicate with some back-end system.
Will edge computing replace cloud computing altogether?
I don’t think they’re mutually exclusive. Edge-based systems deal with some centralized computing resource, and typically that’s going to be the cloud.
“All the edge-based systems I have built in the last four years have been paired with the cloud, because while we’re gathering information independently at the edge, the longer-term, computing-intensive analysis of that data takes place on a back-end system. Ultimately, they’re symbiotic.”
Can you share examples of applications that would need this combination of edge and cloud computing?
They’re going to be industrial. There are thousands of applications where we’re able to place platforms near the devices that they’re looking to control. Some of the more creative ways that I’ve seen it done is in having robots on the factory floor do self-diagnostics and then report back into a centralized system. HVAC systems also can make decisions based on who is in the room without having to communicate constantly with a back-end environment. We’re going to see a lot of edge computing going forward.
What kind of challenges might the move towards edge computing present?
The downside of edge computing is we’re putting another processor, data source, network connection out there—that’s going to be another attack factor that someone can exploit. So, you have to deal with encryption identity, access management, authentication, hardware, even physical security.
“Edge-based computing still needs to evolve a bit in dealing with these various systems. You know, securing everything from a robot, to a wearable, to a tractor.”
How will edge computing change our future computing models overall?
I think we’re going to become more distributed in terms of the physical distribution of devices. We don’t seem to have any way around that based on what we want to do in the future—and what users are demanding with the digital enablement movement. Hopefully, that will be with centralized controls so the devices themselves will be centrally controlled, managed, secured, and will have those capabilities, versus managing them as a one-off (which will drive you nuts if you’re dealing with millions of devices out there under your control.) That seems to be where we’re moving going forward, with edge-based computing as we need it to push some processing close to the sources.
David S. Linthicum is a chief cloud strategy officer at Deloitte Consulting, and an internationally recognized industry expert and thought leader.
This content is produced by WIRED Brand Lab in collaboration with Western Digital Corporation.