Attack on x86 Qualcomm Introduces Arm Processors for Industrial PCs

Related Vendors

NEXCOM International Co., Ltd.

congatec GmbH

Qualcomm, Inc.

ADVANTECH Europe B.V.

FAULHABER Drive Systems

more less

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Wuxi InfiMotion Technology Co., Ltd.

less

Qualcomm enters the industrial PC processor market with the Dragonwing-IQ-X series. The chips combine Oryon CPU, integrated NPU, and Windows 11 IoT support, targeting PLCs, panels, and edge controllers. Initial partners like Advantech and Congatec are preparing modules and box PCs.

With the Dragonwing-IQ-X series, the company introduces its first processors explicitly designed for industrial PCs and edge controllers. At the core is the Oryon CPU architecture, which Qualcomm already uses in its Snapdragon X PC platforms. In the Dragonwing IQ-X, it scales from eight to twelve high-performance cores and relies on a 4-nm manufacturing process. It is complemented by an integrated NPU with up to 45 TOPS of computing power for AI applications directly at the edge.

Oryon CPU And NPU for Harsh Environments

The series is designed for long-term use in harsh industrial environments. Qualcomm specifies an industrial temperature range of −40 to 221 °F, and the package is built to withstand shock, vibration, and limited cooling. This targets control cabinets, machine housings, and edge nodes, where x86 CPUs have dominated so far.

To simplify integration, Dragonwing IQ-X supports common computer-on-module form factors. OEMs can use the new Arm SoCs as a drop-in replacement on existing carrier boards. An evaluation kit is intended to accelerate entry into design-in projects. The platform offers extensive interfaces, including PCI Express, USB, display interfaces, and typical peripheral connections for industrial I/O.

Important for many automation engineers: the SoCs support Windows 11 IoT Enterprise LTSC. Qualcomm cites typical software stacks such as Qt for visualization, Codesys and EtherCAT for control and fieldbus connectivity, as well as proprietary AI tools based on the Qualcomm AI Software Stack. Models can be deployed to the NPU via common runtime environments like Onnx and PyTorch. This brings applications such as condition monitoring, predictive maintenance, or automated visual inspection directly into the industrial PC.

OEM Partners Plan Modules And Box PCs

For market entry, Qualcomm is relying on well-known industrial PC providers. The first partners include Advantech, Congatec, Nexcom, Portwell, Seco, and Tria. They are working on computer-on-module designs and compact box PCs that will integrate Dragonwing IQ-X into control and visualization tasks. Initial products are expected to follow in the coming months, some in the COM-HPC Mini format.

An example is provided by Tria with the HMM-IQX module: it combines a Dragonwing-IQ-X variant with up to twelve Oryon cores, an Adreno GPU, and the Hexagon NPU. The module offers up to 64 GB of LPDDR5X memory, multiple USB4 and PCIe links, as well as DisplayPorts. Such components are targeted at compact control computers, AI-supported image processing, or edge gateways that locally preprocess sensor and machine data.

Architecture Shift in Industrial PCs

With Dragonwing IQ-X, the pressure on established x86 providers in the industrial PC segment increases. So far, the segment has been dominated by a mix of classic x86 CPUs and external GPUs or accelerators when AI or image processing is required. Qualcomm aims to combine CPU, GPU, and NPU in a single Arm SoC, thereby reducing power consumption, component count, and cooling effort.

Whether the step is successful depends heavily on the ecosystem. Long-term availability, real-time capability, deterministic communication, and support for existing automation software are more important to machine builders than raw benchmarks. With Windows 11 IoT support, COM modules, and partners from the IPC world, Qualcomm is at least addressing the key levers. For the automation industry, this signals another attempt to establish Arm-based platforms in control technology, this time with an integrated NPU and a clear focus on edge AI. (mc)

Machine Authentication

Machine Operator Authentication - Why knowing "who" matters