It wasn’t that long ago when with servers in our offices, we all had edge computing. Of course, we didn’t think of it that way. It was just what worked. Then, along came the cloud, and everything changed. Computers were hundreds of miles and milliseconds away. For our office applications, that level of latency’s OK. But with the rise of the Internet of Things (IoT), 5G, and our never-satisfied need for speed, a new kind of local computing, edge computing, is appearing.
And, it’s not just showing up as another technology du jour. Arpit Joshipura, The Linux Foundation‘s general manager of networking, predicts, “edge computing will overtake cloud computing” by 2025.
IBM Services Global CTO and Vice President Bridget Karlin, won’t go quite as far as Joshipura, but she believes, “We will see an increase in Edge computing due to the sheer quantity of instances compared to centralized cloud centers. IBM estimates that there are some 15 billion intelligent devices in the market today, and IDC forecasts that by 2025 that will grow to 150 billion — resulting in unprecedented volumes of data.”
Lewis Carr, senior director of product marketing and management for Actian, a cloud data management company, can also see edge computing dominating. Carr thinks the “edge will overtake cloud in terms of sheer horsepower, data collected, and even number of cycles on data processing and analytics operations applied to that data locally at the edge — provided we take the edge to mean end-to-end across the various tiers of the edge.”
What he means by that, is that “you can’t take the edge as just the complex IoT devices, such as smartphones and application-specific dedicated machine vision systems. Instead, you must include the lower-end grids of IoT devices with 8-, 16- and 32-bit microcontroller units, as well as the remote and field servers and mini data centers supporting such environments as transoceanic cargo ships, forward command posts, rural hospitals and so forth.” In short, edge computing will range from our homes to our offices and to our factories. It will be everywhere.
When Joshipura is talking edge computing, he has a very specific meaning in mind. It’s compute and storage resources, which are only five to 20 milliseconds away.
What needs such small latency — besides hyperactive Fornite gamers? Our new generation of hardware, which is all about working in real-time and the applications running on top of them. This includes manufacturing equipment monitoring, self-driving cars, telemedicine, and a host of devices and applications we’re still inventing.
Individually, each device’s network traffic isn’t that much. But, with thousands of, say, video cameras being used for videoconferences, telemedicine visits with five-year-old childern, and CCTV security monitoring, high latency and bandwidth costs can cripple these applications.
Related: Edge computing and AI technologies: Turning data into instant insights
There are also use cases where edge computing power enables more successful applications. Bob Gill, a Gartner research VP has pointed out, for example, “Let’s say there are a whole series of traffic lights in a city, with edge computing one of the things we can do is actually provide intelligence between devices, so they can communicate between themselves without having to talk to a central locale.”
Thanks to the edge’s local processing and storage power, these new applications can work properly. Sure, some data still needs to go to the cloud or your remote data center, but less does.
Open-edge computing
From Joshipura’s perspective, the best way forward for edge computing is with an open, interoperable framework. This framework should be independent of hardware, silicon, cloud, or operating system. Open-edge computing should also work with any edge-computing use case: IoT edge, telecom edge, cloud edge, or enterprise edge, whatever, the “goal here is to unify all of these.”
One way this is being done is with the LF Edge. This Linux Foundation organization seeks to bring all edge computing players under one umbrella. Its job is to create a software stack that unifies a fragmented edge market around a common, open vision,
That all sounds well and good for edge computing users and companies, but why does Joshipura think that edge computing will overtake and surpass cloud computing in just five years? After all, Gartner estimated the total worth of the public cloud market in 2019 was $214.3 billion, with a growth rate of 17.5%.
Joshipura argues that edge computing potential has been hidden by too many cooks spoiling the broth. Too many previous edge computing efforts have been stuck in narrow silos or have worked at cross purposes. Instead, by focusing on the communalities and getting everyone on the same page, he believes edge computing has tremendous enormous potential.
Not everyone buys that Joshipura’s LF Edge’s open approach will win out. Carr said, “I’m not sure his nirvana around a single Linux platform independent of everything else will come to pass — edge has always been fragmented, but I do see it narrowing down to the vast majority being two platforms for each layer: At the CPU level, Intel and Arm; two OSes Embedded Linux (with variations) and Android; and two private and public cloud architectures where vendors must support all options at each of these levels.”
Where does this new edge computing lead us?
Bridget Karlin, IBM Services Global CTO and Vice President, sees 5G and Wi-Fi 6 as empowering edge computing for many new roles. In particular Karlin sees it enabling autonomous cars.
For example, the automotive outlook for 5G involves uber-connectivity, with networked sensors inside and outside of the car, monitoring everything. Autonomous vehicles are becoming edge compute instances that will process a range of analytics across a wide variety of applications. They are ingesting the data from sensors, performing the AI-based analytics, and triggering real-time instructions, all happening in the vehicle. Additionally, enhanced safety, awareness, and infotainment will all be part of the 5G connected car. This is edge computing: bringing the compute to the vehicle, close to where the data originates. The 5G standard will connect cars with surroundings, other drivers, and infrastructure in ways that we can’t comprehend yet.”
Julius Francis, Juniper Networks‘ director of product management and marketing, agrees that the edge will flourish with the intersection of 5G and IoT. Francis lists the following in his vision of 2025 edge computing:
- Applications driving video and mobile traffic will continue to grow with the 5G rollouts and demand the edge computing for higher bandwidth and lower latency experience.
- The emerging applications such as high-frequency trading, autonomous cars, AR/VR and multi-player cloud gaming are also going to drive more and more workloads at the edge.
Further, Francis concluded, “Once the high immersive edge computing power is unleashed, we will see applications that are not yet imagined coming to the edge. These scenarios are going to shift the edge adoption quicker than centralized cloud adoption.”
To turn this from science-fiction into science fact, Karlin believes “Modernizing with a highly elastic and extensible multi-cloud infrastructure to unify the capabilities within an entirely governed team-platform is the key. With this in place, enterprises are poised to lead in the 5G and edge generation, and those companies that take steps now to prepare their hybrid, multi-cloud strategy, infused with AI-based insights, will be ready to capitalize on its full potential.” Still, “while edge computing is the future, cloud computing is and will always be a critical component of that.”
Related: A fight is brewing over edge computing
Other edge computing takes
Not everyone sees edge computing as a bridge between local devices and services. For example, Jay Valentine, VP of operations for Cloud-Sliver, argues that while “the majors like Dell, HP, IBM, Nutanix are all trying to make their machines smaller and put them at the edge, nothing changes. Stuff runs the same as it did in the cloud or data center, just closer to the customer.” Valentine argues, heavy server-based applications that require a million dollars to operate, can be run instead “on a Raspberry Pi that costs less than $50.” That certainly may be true of some applications, but it’s hard to see this micro-device-based approach as becoming a major edge computing trend.
For an entirely different alternative take, Larry Aultman, Founder and CEO of Intact, a company bringing legacy applications to the edge, said he “fundamentally disagrees with the ‘overtake cloud computing by 2025’ statement. This is the latest in the never-ending shell game of ‘re-name’ that tune! This is just another attempt to redefine what the word ‘edge’ means or at least bend it to fit your world view. There is no consensus on what Edge means. Grid, edge, distributed (whatever you like to call it) is only just emerging as a commonly accessible platform with the introduction of 5G networks. It will take at least another 8 years for the biggest consumers of cloud services to bring true ‘edge’ computing into their stack.”
The edge and the cloud
Perhaps the wisest take on the future of the edge and the cloud comes from Vibhoosh Gupta, product management leader for Emerson’s machine automation solutions business. “They are more complementary than competitive. We believe future systems will include both cloud and edge components.”
While the term “edge computing” has been used for many different technology approaches, it seems clear that this new sub-20-millisecond vision will be the one that shapes our future.
I don’t know if edge computing will outpace the cloud in terms of capital investment, CPUs, or revenue. What I do know is that, like such fundamental technology shifts before it, like the cloud, the internet itself, and the web, edge computing will create a shift in how we use computing both in our lives and daily work.
