Edge computing is getting “edgier.” 

The technology, which puts compute and processing resources closer to where data is being analyzed, is moving farther and farther away from on-premises data centers—and it’s moving fast. By 2025, IT research company Gartner predicts that more than 50% of enterprise-managed data will be created and processed outside the data center or cloud.

As Pure Storage® was just named one of The 50 Hottest Edge Hardware, Software and Services Companies of 2022 by CRN, I polled colleagues and Pure’s partners for the trends in the state and spread of edge computing for the next year. Here’s what they came up with.

Sign up for email1. Edge computing will be all around—and down on the farm. 

What’s the “edge” of a farm? It’s animals, tractors, soil, and more. If you can attach a sensor to something in the ag environment, it’s definitely an edge.

Upgraded edge computing capabilities open up the technology for more industries—those in far-reaching places, difficult to access terrain, or simply dangerous conditions. (It’s one of the main perks of automation, cobots, and IoT—keeping humans safe and out of harm’s way.)

Agriculture shows some of the most promise for edge computing. Farmers use edge technology to track water use and animals, decide where to put fertilizer and in what amounts, analyze soil quality, and monitor crop growth. Even tractors can become part of an edge network, along with sensors spanning fields.

2. Filtering edge data will deliver a competitive edge.

The smarter edge computing gets, the more it may be able to filter out the noise that a central hub would have to process and sift through for value. As my colleague said, “It’s like delivering the CliffsNotes of telemetry data.” 

For the Mercedes-AMG Petronas Formula 1 team, this type of edge can deliver competitive edge. Hundreds of sensors on each car provide real-time data about performance—terabytes of data per race. But what about the effort to filter out less valuable telemetry? The team has plans to use smarter sensors that do more processing at the edge, putting more valuable data in the hands of engineers.

It’s an example of optimization and performance monitoring that can be game-changing for F1 but also for numerous other business breakthroughs.

3. Edge computing doesn’t need to be connected. 

Sure, “connected” is good because the data gets back to centralized data centers faster. But in many ways, only being able to deploy edge systems in a connected fashion was not only difficult but also made certain offerings unavailable—especially to more remote industries and operations. 

What many organizations don’t realize is that edge computing data operations can also operate in a disconnected mode. An example: an edge system running autonomously in a hard-to-reach place like a rural mining site. The system will continue running when a connection drops. Then, when the connection resumes, the system can sync and transfer data without disruption to the business. This disconnected mode opens up possibilities for much more.

Whether as part of the core design of an edge system or a function of resiliency, disconnected edge just might be the next big thing.

4. 5G infrastructure will knit together internet of things (IoT).

IoT edge computing is accelerating the next generation of automation and is one of the core components of the industrial internet of things (IIoT), where industries leverage edge platforms for analytics, smart buildings, and more. To keep the mesh of wired and wireless devices from unraveling, ultralow latency 5G networks will deliver constant connectivity that’s fast and reliable. 

Take a manufacturing sensor in a factory, for example. The sensors and actuators connected to the machinery (the edge) would create a pod, and a message broker orchestrates communication of telemetry data between the sensors and data processing service. The data is ingested and stored in a stateful microservice that needs persistent storage before moving to the cloud to train ML models. Back at the edge, trained models can detect anomalies on the floor and predict maintenance of equipment. 

Learn more about running AI and IoT workloads at the edge.

5. The edge is getting “foggy.” 

In this entire conversation, we’re talking about data that’s being generated off-site, as opposed to how and where you’re doing things with that data. Conventional edge computing is about parsing out the useless data, so that all you have is the useful data. 

Fog computing, as it’s known, is sort of the “edge of the edge.” The concept is similar to edge computing in that it brings processing to the edge—but it goes farther. It doesn’t rely on the cloud but instead carries out much more computation, storage, and communication locally at the edge—like “micro” data centers that are purpose-built with power, monitoring, physical security, cooling, and interconnections that can be deployed.

6. Containers (and container-native data storage) will be key.

Containers and Kubernetes make an ideal platform for the edge. Hyperscale cloud providers are taking note with AWS Snowball and Greengrass for IoT, Azure Stack, and Google Anthos, which are all based on Kubernetes. Anthos is intended to be a foundation for 5G infrastructures and data analytics and AI workloads, while all of the offerings run data ingestion, data storage, data processing, and machine learning workloads at the edge. 

But what about the challenges of running data-centric workloads at the edge? Containers and container-native storage can provide core services, persistent storage, high availability, and durability. It also can enable seamless migration between the cloud and edge with minimal effort.

Learn more about edge computing with containers and how Portworx® can help.

7. The edge will meet Web 3.0.

As the edge expands, how will challenges with efficiency and latency be addressed? “Decentralization,” a cornerstone principle of Web 3.0 technologies and the blockchain, will come into play. Multiple servers spread across networks can remove central points of failure and allow edge environments to tap a broader network of compute power.

8. As a service will accelerate edge tech, too. 

Emerging software-defined networking and network-as-a-service companies could transform (and accelerate) edge networking capabilities and adoption. This means delivering private, next-generation networking as a service that can replace MPLS and SD-WAN networks and speed adoption. This will also make it easier to scale, but security should be built in from the ground up. And there will be additional benefits, such as automated edge network management.

The leaders in this networking space will be the providers who can unify disparate edge IT components into a central network, giving customers true control and visibility into their own edge networks.

9. Chips and operating systems are making edge devices smarter and more powerful.

Edge devices are the sum of a lot of powerful, cutting-edge technologies, including chips, operating systems, processors, batteries, and more. They’re all optimized for the edge, with power distribution and management in mind. This helps keep edge apps and devices running—and prevents data loss that occurs when they don’t.

The trend toward powerful and smart edge devices is driven by the increase in edge-optimized operating systems, including IGEL OS, a Linux operations system that’s built for enterprise access to virtual environments. The trend is also accelerating thanks to powerful chips like Xeon Scalable chips from Intel, which transfer data rapidly in edge servers. 

How Pure Storage and Our Partners Are Leading the Way in Edge Computing

We’re proud that Pure was named to CRN’s “2022 Edge Computing 100” list. As CRN noted, “Pure Storage’s FlashBlade is a unified fast file and object platform that can be used with AI workloads and heavy data processing requirements at the edge. FlashArray, meanwhile, puts VMware inside Pure Storage’s all-flash storage to create a hybrid cloud product that supports a 5G multimode infrastructure.” 

Our partners and MSPs play a key role in providing our edge computing solutions, and many of them also made the CRN list: 

  • Equinix provides infrastructure and services to help manage data, including Network Edge services, which is optimized for deployment and interconnection of network services.
  • NVIDIA enables large, complicated workloads at the edge with NVIDIA AI-on-5G, a unified platform for deploying AI-dependent applications over a mobile network.
  • Splunk is an analytics powerhouse, and Splunk Edge, specifically, can accept sensor data without configuration. They also offer hardware, with Splunk Edge Hub that has built-in sensors that can relay data back to a Splunk Cloud platform.
  • Veeam has an edge-specific backup offering for edge environments on Azure Stack Edge and Kasten Kubernetes backup and disaster recovery for containers in production.

Be ready for whatever comes your way in 2023 and beyond with the most complete Kubernetes data services platform. Learn more about Portworx.