10 edge computing trends to watch in 2026 and beyond

1. AI ambitions drive increasing adoption of edge computing

Enterprise needs and objectives have always driven the demand for computing capability at the edge

But those needs and objectives have changed in recent years.

Up until a couple of years ago, the key issues driving enterprise adoption of edge computing were cost and bandwidth efficiency, data privacy and security, business continuity and resilience, sustainability and energy efficiency, and vertical-specific use case customization.

Those factors are still in place, but increasingly, an organization's AI ambition is the primary reason for its edge computing needs.

"Organizations are trying to operationalize AI and real time decision-making, and that makes compute locality, resilience and governance core architectural requirements. Done well, edge improves experience, accelerates decisions and helps meet regulatory and data-sovereignty obligations," said Titus M, practice director at research firm Everest Group and leader of its cloud and infrastructure services research practice. "[So edge computing is] now less about baseline digitization and more about operationalizing AI plus real time controls and meeting data residency and locality [needs]."

McKinsey & Co. also noted this dynamic, writing that "while massive, centralized clusters still dominate the landscape, businesses are training and deploying AI models across multiple locations, including edge environments, to optimize performance, reduce latency, and improve resource availability. This distributed approach not only alleviates computational bottlenecks but also enhances resilience by reducing reliance on concentrated infrastructure."

2. The designs and capabilities of edge devices are improving

The quality of edge computing devices is also on the upswing, said Brian Alletto, director of cloud technology at digital services firm West Monroe.

He said there is an increased "robustness of construction" as the technology incrementally advances each year. Vendors have made the edge devices more flexible so they can easily be mounted or configured in different ways to fit the use case requirements and the locations in which they're placed. They have more insulation to protect them from the elements in the environments where they operate, such as extreme temperatures and vibrations. Their computing capabilities and power efficiency have also increased.

Additionally, vendors have introduced better management software platforms and frameworks, Alletto said. This enables organizations to more easily deploy, manage, secure, and maintain edge devices at scale, "away from the traditional IoT centers of excellence."

All these improvements permit organizations "to deploy edge technologies in ways that are cost-effective, durable, reliable and suitable for deployments outside a nicely designed and protected data center," Alletto added.

3. Spending on edge technology will continue to soar

Market value figures vary widely. But there's consensus across multiple research and analyst reports that spending on edge is rising and will continue to grow.

According to research firm IDC's May 2025 market forecast, global spending on edge computing was expected to reach nearly $261 billion in 2025 and grow at a compound annual growth rate of 13.8%, reaching $380 billion by 2028.

Figures from Global Market Insights, while lower overall, also show significant growth. The firm estimates the global edge computing market at $21.4 billion in 2025, growing to $28.5 billion in 2026 and $263.8 billion in 2035, at a compound annual growth rate (CAGR) of 28%.

Meanwhile, Precedence Research calculated the global edge computing market size at $554.39 billion in 2025. It predicted the market will reach $709.91 billion in 2026 and exceed $6 trillion by 2035, growing at a CAGR of about 27% from 2026 to 2035.

4. Edge options continue to expand

The number and type of edge computing devices and deployments are expanding.

"Edge is a family of technology that includes hardware, software, data and services and ensuring that those elements are located where they can be optimized. And it's becoming much more strategic and broader in nature," said Michele Pelino, a vice president and principal analyst at Forrester Research and leader of the firm's edge computing and IoT research.

Organizations across industries are deploying purpose-built edge computing devices within their own facilities. But that's only a fraction of the edge compute power out in the world today, Pelino said.

Some companies are building second- and third-tier data centers to host edge computing. For example, such data centers might be housed in a commercial building to process data within that specific facility.

Additionally, organizations use content delivery networks, a collection of geographically distributed yet interconnected servers that cache content locally to speed delivery to end users.

Some organizations are buying edge computing capabilities from telecom service providers, whose extensive infrastructure and reach enable them to place edge devices physically close to nearly all potential customers. The telecom operators use that proximity to offer edge computing equipment, services and supporting components, such as secure access service edge, which bundles and delivers network and security-as-a-service functions as a single service.

Even the cloud hyperscalers are offering edge computing options, Alletto said. One example is AWS Local Zones, which AWS said offers "single-digit millisecond latency or local data processing by bringing AWS infrastructure closer to your end users and business centers."

5. More strategic use of the different edge types

Forrester Research categorizes edge computing into four categories: enterprise edge, operations edge, engagement edge and provider edge.

Pelino said these distinctions matter because the technology stacks differ, meaning one is better suited to certain use cases than the others.

For example, the enterprise edge includes remote office data centers and micro data centers to serve organizational needs. The provider edge, in contrast, comprises computing resources accessed using the internet offered by telecom companies, service providers or content delivery networks and is typically for delivering services such as online gaming and content streaming.

"Given the diversity of those environments, you have to think about the technology stack needed to enable the different types of use cases," Pelino said.

She said the enterprise edge, for example, powers smart buildings or applications in a campus-type environment, whereas the operational edge is suited to industries such as healthcare.

Pelino said CIOs and enterprise IT departments may not use these four terms specifically, but they do have an increasing recognition that "there's not just one edge. And it's not just edge. It's usually edge to cloud; they understand that some data needs to go back to the cloud."

This more nuanced understanding of the edge enables enterprise IT teams to design the right edge program for the business problem or process they're trying to address, Pelino added.

6. Edge growth creates infrastructure challenges

The distributed nature of edge computing presents challenges, and its effects are growing alongside the demand for and deployment of edge hardware.

M said there are two persistent challenges here.

"One, scalability of edge infrastructure continues to be a challenge from a cost economics point of view (different devices for compute, analytics, etc., add more cost and lower ROI). To date, we haven't seen disruptive offerings that can help edge computing provide similar scalability that the cloud does," he said. "Second, power requirements, cooling and physical security are additional challenges we have seen."

Others cited additional challenges and issues.

For example, the sustainable management of edge assets across multiple locations and devices can be a significant concern for some organizations.

"The location of the equipment will be on pedestals and roadside vaults, and that creates challenges with powering and cooling them," said David Witkowski, a senior member of the professional association IEEE and co-chair of the Deployment Working Group at IEEE Future Networks.

The overall risk of using public spaces to house edge equipment is another concern.

"It creates challenges with the security of the vaults [that house them]. These assets are located in public places that could be compromised, and even getting space to place them will be challenging for companies deploying this equipment," said Witkowski, who is also CEO of Oku Solutions, which provides professional support services to the wireless telecommunications industry.

While telecom service providers and others building up edge computing capabilities are working through such issues, these challenges could slow down the pace of edge deployments and slow the speed of innovation and the use cases -- such as self-driving vehicles -- that rely on the technology to operate, he said.

There are advancements on this front, however, Alletto said. Improvements in both management software and edge deployment best practices are making it easier for more organizations to remotely monitor their edge devices, often with fewer people.

7. More hackers are targeting edge deployments

Threat actors have noted the growing number of IoT and edge computing devices as prime targets.

At the same time, organizations face challenges securing what, for many, are complex and expansive edge environments.

"In contrast to traditional cloud environments, where data processing takes place in centralized, secure data centers, edge computing involves multiple endpoints that are frequently deployed in remote or uncontrolled locations. The decentralization of edge computing creates a greater number of potential entry points for cyber-attacks," explained Allianz's 2025 report.

The report further noted that "the heterogeneity of devices used in edge computing systems exacerbates security vulnerabilities. Edge environments often include a diverse array of devices from various manufacturers, each with its own operating system, firmware, and communication protocols."

Researchers have identified numerous potential threats, including the following:

• Attacks against user and endpoint devices.
• Sniffing attacks against the radio access network (RAN).
• Attacks against servers and data at the network edge.
• Sniffing attacks against endpoint (user) devices and components.
• Attacks against associated cloud workloads.
• Attacks against applications at the network edge.
• Supply chain attacks.
• Attacks against the 5G core network.
• Physical attacks against technical components such as IoT devices and abandoned assets.
• DDoS against RAN.
• Attacks against multi-access edge computing.

"We're talking about fragmented connected devices that open up doors to more bad things happening," Pelino said. "Companies need to think about security upfront and very early on."

8. Edge AI is poised for substantial growth

That higher computing power is enabling AI and machine learning to move from the cloud to the edge, and the ability to support AI with edge computing continues to improve.

"AI-specific edge chips, such as Nvidia's Jetson series, bring unprecedented compute power to the edge, enabling sophisticated AI inference on smaller, energy-efficient devices," M said. "This opens the door for a wide range of use cases, from AI-driven retail kiosks to predictive maintenance in industrial settings."

The developments in AI-capable edge chips are not only accelerating innovation, M said, but also influencing broader ecosystem trends. Enterprises are rethinking their IT architectures to prioritize hybrid models that blend edge and cloud capabilities, using the strengths of both.

"Additionally, edge-specific AI frameworks and platforms are becoming more prevalent, making it easier for developers to deploy and scale edge AI applications," M said.

Additionally, M said AI at the edge is "moving beyond classic [machine learning] into multimodal and generative inference, and that's pushing a wave of inference-optimized silicon and software stacks. The winning pattern is smaller, optimized models running locally for latency, cost and privacy, while the cloud remains essential for training, coordination and lifecycle management."

M further noted that the market is "seeing more complete edge MLOps capabilities that make it feasible to operate thousands of AI endpoints consistently."

9. 'Powerful AI can happen at the AI PC level'

AI PCs are a new class of computers that include a neural processing unit (NPU) alongside a conventional CPU and GPU, enabling them to run AI tasks locally.

"With AI PCs, powerful AI can happen at the PC level," Pelino said. "So now you can do those powerful analytics on a laptop in remote locations, whereas before you had to use a data center."

Pelino said that as more users get AI PCs, it will bring "edge to life in a variety of different environments we hadn't seen in the past."

10. As edge computing matures, use cases for edge computing will grow

As edge computing matures and becomes more powerful, it can support a larger, more diverse set of use cases, Pelino said.

"It is really opening up many new use cases," she added.

For example, organizations can deploy more reliable compute power at the edge in remote locations, thereby allowing them to expand operations or address problems they couldn't before, Pelino said. Think about adding robotic capabilities in more places, thanks to real time intelligence capabilities delivered with low latency at the edge.

Mary K. Pratt is an award-winning freelance journalist with a focus on covering enterprise IT and cybersecurity management.