Back to All
OnQ Blog

Edge AI ignites the next industrial revolution

Scaling power-efficient AI at the industrial edge with Qualcomm Dragonwing to accelerate digital transformation
Qualcomm-image



What you should know:
  • Edge AI is the catalyst accelerating the convergence of operational and information technologies (OT & IT), transforming data into real-time insights. This enables faster decisions, more predictable costs, enhanced privacy and greater operational autonomy.
  • Scaling AI across industrial environments remains a challenge, with data sovereignty and hardware limitations often standing in the way of efficient deployment. Purpose-built edge AI platforms are essential for moving from prototype to production at scale.
  • Qualcomm Dragonwing IQ Family of edge AI processors and AI on-prem appliance are designed to optimize industrial applications, delivering the performance, efficiency and scalability needed to turn transformation goals into measurable business outcomes.



We are living through a fundamental shift in how intelligence works. Artificial intelligence is woven into the fabric of the world around us, embedded in the devices we use, the infrastructure we depend on and the systems that keep economies running. Our vision is an AI compute continuum, where cloud and edge AI work seamlessly together to deliver the speed, scale and intelligence required in an increasingly connected world. This is the age of ambient intelligence, where systems can perceive, reason and act in real time.

Nowhere is this transformation more consequential than in the industrial verticals that form the backbone of the global economy. Manufacturing, oil & gas, energy & utilities, logistics and more are undergoing profound reinventions as AI accelerates the convergence of operational technology (OT) and information technology (IT). Data that was once isolated in machines, sensors and control systems can now be transformed into real-time intelligence, connecting physical processes with agentic AI workflows and autonomous decision-making, helping organizations improve productivity, optimize operations and respond faster to ever-evolving business demands.

Qualcomm-image
Figure 1: AI is transforming a wide range of industrial segments.

Edge AI is essential for the future of industries

The old model of centralized control, manual inspection and reactive decision-making is giving way to a new paradigm defined by distributed intelligence, predictive insight and autonomous action. Industrial operations that once relied on human observation and periodic reporting now demand continuous awareness and intelligence across every asset, process and site.

But to efficiently scale, AI needs to move to the OT edge. Edge AI enables intelligence where it matters most — directly on the machines or devices, at the moment of action. By processing data locally rather than sending it to the cloud, edge AI eliminates latency, reduces bandwidth costs and keeps systems running even when connectivity is unavailable. For industrial operators, this translates into faster decisions, fewer disruptions and lower operating costs. And because data stays on-premises, organizations maintain full control over sensitive operational information — a critical advantage in an environment where security and compliance are non-negotiable. It is a strategic imperative that gives industrial organizations full control over their data, their infrastructure and their competitive advantage.

Edge AI is making this shift possible today, not as a future aspiration, but as an operational reality deployed in factories, substations, pipelines, warehouses and fields around the world. For industrial leaders, the question is no longer whether AI will reshape their industry, but how quickly they can scale it and how deeply they can embed it into the core of their operations.

Qualcomm-image
Figure 2: Edge AI accelerates the convergence of IT and OT.

We are establishing the foundation for industrial edge AI

To help realize the future of industry, we have built a comprehensive portfolio of industrial connectivity and edge AI solutions that bring powerful, efficient and reliable intelligence directly to industrial environments. From the Dragonwing IQ Family of edge AI processors that are engineered to power a diverse spectrum of industrial applications to our Dragonwing AI on-prem appliance, we offer a scalable portfolio of industrial solutions that meet their most stringent requirements. Together, these solutions form part of the broader Dragonwing portfolio, bringing AI, high-performance low-power computing and advanced connectivity together to help industries accelerate digital transformation.

Let me spotlight a few key offerings in our edge AI portfolio and explain how they are scalable to the needs of industrial AI adoption.

 

Building a scalable industrial edge AI infrastructure

Whether the requirement is high-performance AI for advanced computer vision, robotics, generative AI, automation and multi-sensor fusion, or ultra-efficient inference for always-on embedded systems operating within tight power and thermal budgets, the Qualcomm Dragonwing IQ Family provides a scalable platform optimized for industrial AI deployments. It allows industrial OEMs, system integrators and solution providers to deploy the right level of AI across a wide range of performance, power and cost requirements, so that they can confidently scale ideas — from pilot projects to production-ready systems — across entire fleets, facilities and operations.

At the high end, the Dragonwing IQ9 Series delivers up to 100 TOPS of AI performance for demanding workloads in edge AI gateways, autonomous mobile robots and industrial automation systems. Designed for industrial environments, it combines high-performance computing with power efficiency, available built-in safety features, design flexibility, Linux Yocto and Ubuntu support, and an extended operating temperature range. The IQ9 is available today and it can be evaluated with the IQ-9075 EVK.

For applications that prioritize power efficiency without compromising AI capability, the Dragonwing IQ8 Series provides up to 40 TOPS of AI performance. It is well suited for applications such as fleet management systems and healthcare hubs, while offering industrial-grade software support and extended temperature range. The platform is available through the IQ-8275 EVK today and powers the new Arduino VENTUNO Q.

At a tier below, the Dragonwing IQ6 Series delivers 1.1 TOPS of AI performance for applications including HMIs and industrial controllers, bringing AI capabilities to cost- and power-sensitive deployments.

Beyond what’s available today, we continue to expand the portfolio across the performance spectrum, from the recently announced Dragonwing IQ10 Series for advanced robotics to new offerings that will further extend power-efficient AI capabilities across the industrial edge. On this, something new is coming on the horizon, so stay tuned.

The Qualcomm Dragonwing portfolio is purpose-built to deliver scalable, power-efficient industrial AI performance, with the Dragonwing IQ Family providing the edge AI foundation for a wide range of industrial deployments.

Qualcomm-image
Figure 3: Dragonwing IQ family of edge AI processors.

Redefining industrial PCs for challenging environments

The Dragonwing IQ-X brings the full power of the IQ Family into the industrial PC form factor — making it the processor of choice for operators who need workstation-class AI compute in a ruggedized, facility-deployable package.

The IQ-X targets industrial PCs, HMIs, single-board computers and automation systems that require a familiar PC formfactor in demanding industrial environments. Qualcomm Technologies collaborates with Canonical to provide certified Ubuntu images and with Microsoft to support Windows 11 IoT Enterprise LTSC, including certified drivers, long-term OS support and regular security updates. By integrating CPU, GPU and NPU acceleration into a single SoC, the IQ-X brings AI directly into industrial edge without requiring changes to existing tools, workflows or software ecosystems. These collaborations help reduce deployment complexity and accelerate time-to-value for customers building next-generation industrial solutions.

Qualcomm-image
Figure 4: Dragonwing IQ-X powers next-generation industrial compute.

Bringing cloud-grade AI to on-premises deployments

For deployments that require even greater AI performance at the edge, the Dragonwing AI On-Prem Appliance delivers up to 870 TOPS of AI compute in a compact 150W system. Built on Qualcomm Cloud AI 100 Ultra accelerators and available through partners including Aetina, Advantech and Lanner, it supports large vision and language models up to 120 billion parameters, multi-user inference and RAG workloads. This enables video analytics, forensic search, natural-language interaction and operational intelligence to run locally, helping organizations meet privacy, compliance, latency and cost requirements while keeping data on-premises.

Qualcomm-image
Figure 5: Dragonwing AI on-prem appliance.

Lastly… Dragonwing is built to last

In industrial environments, hardware is built to run for a decade or more, making product longevity support a fundamental requirement, not an optional feature. When a processor reaches end-of-life mid-deployment, manufacturers face costly redesigns, recertification processes and supply disruptions that ripple across their entire product line. Qualcomm Dragonwing IQ Family addresses this directly with hardware and software support extending to 10 years or more, giving industrial OEMs and system integrators a stable, long-lived foundation they can build on with confidence.

Qualcomm-image
Figure 6: Product longevity program for Qualcomm IoT portfolio.

Our vision for the future

The industrial world stands at an inflection point. As intelligence scales across devices, networks and infrastructure, industrial edge deployments will become an increasingly important part of a larger AI ecosystem spanning the compute continuum. Edge AI is an active transformation already underway in factories, energy grids, logistics networks and cities across the globe. Looking ahead, the full potential of this shift is still ahead of us: a future where every machine is aware, every process is self-optimizing and every decision is informed by real-time intelligence generated at the source.

For industries, this means unlocking levels of productivity, safety and sustainability that were simply out of reach before. For society, it means cleaner energy systems, safer workplaces, more reliable infrastructure and technology that extends opportunity to communities that connectivity alone could never fully serve. The companies and institutions that lean into edge AI today are not just improving their operations — they are helping to define what the next era of industrial civilization looks like. That future is being built now, one intelligent edge device at a time.

So, take a closer look of our Dragonwing IQ Family EVKs and start today.




Go Deeper
Which Dragonwing processor tier is the right fit for my application?

The Dragonwing IQ Family is designed to scale across the full spectrum of industrial requirements — from the IQ6 Series at 1.1 TOPS for cost- and power-sensitive deployments like HMIs and industrial controllers, to the IQ8 Series at 40 TOPS for fleet management and healthcare hubs, all the way to the IQ9 Series at 100 TOPS for demanding edge AI gateways, autonomous mobile robots, and industrial automation systems. For even greater on-premises AI performance, the Dragonwing AI On-Prem Appliance delivers up to 870 TOPS in a compact 150W system.

How do I start evaluating Dragonwing edge AI processors for my project?

We make it straightforward to move from concept to working prototype — the IQ9 Series is available today through the IQ-9075 EVK, and the IQ8 Series is available through the IQ-8275 EVK, so your team can begin hands-on evaluation against your specific performance, power and cost requirements without delay.

What's next for the Dragonwing portfolio — and how do I stay ahead of it?

We continue to expand across the performance spectrum, including the recently announced Dragonwing IQ10 Series for advanced robotics — and something new is already on the horizon. The best way to stay ahead of what we're building next is to plug directly into our IoT ecosystem updates.

Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries ("Qualcomm"). The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.

Qualcomm branded products are products of Qualcomm Technologies, Inc. and/or its subsidiaries.

About the Author
Danny Tseng
Danny TsengSr. Director, Product Marketing, Qualcomm Technologies, Inc.

© Qualcomm Technologies, Inc. and/or its affiliated companies.

Snapdragon and Qualcomm branded products are products of Qualcomm Technologies, Inc. and/or its subsidiaries. Qualcomm patented technologies are licensed by Qualcomm Incorporated.

Note: Certain services and materials may require you to accept additional terms and conditions before accessing or using those items.

References to "Qualcomm" may mean Qualcomm Incorporated, or subsidiaries or business units within the Qualcomm corporate structure, as applicable.

Qualcomm Incorporated includes our licensing business, QTL, and the vast majority of our patent portfolio. Qualcomm Technologies, Inc., a subsidiary of Qualcomm Incorporated, operates, along with its subsidiaries, substantially all of our engineering, research and development functions, and substantially all of our products and services businesses, including our QCT semiconductor business.

Materials that are as of a specific date, including but not limited to press releases, presentations, blog posts and webcasts, may have been superseded by subsequent events or disclosures.

Nothing in these materials is an offer to sell or license any of the services or materials referenced herein.