CES 2026: Embedded and IoT Qualcomm Unveils New IoT Chips for On-Device AI

From Manuel Christa | Translated by AI 2 min Reading Time

Related Vendors

Qualcomm consolidates its IoT acquisition series and unveils new Dragonwing-Q processors at CES 2026. The package is designed to help developers, OEMs, and major users bring edge AI from prototype to rollout faster.

Qualcomm IoT ecosystem: Dragonwing processors form the hardware base, topped by a unified software stack for Linux, Windows, and Android, along with services for secure device management, on-device AI, video analysis, and positioning.(Image: Qualcomm)
Qualcomm IoT ecosystem: Dragonwing processors form the hardware base, topped by a unified software stack for Linux, Windows, and Android, along with services for secure device management, on-device AI, video analysis, and positioning.
(Image: Qualcomm)

The CES 2026 in Las Vegas marks a stage for a strategic shift for Qualcomm: moving away from pure chip sales toward a comprehensive toolkit of processors, software, and services. The company speaks of a "completed" expansion of its industrial and embedded IoT business. This is based on new Dragonwing-Q processors and several acquisitions over the past 18 months, including Arduino, Edge Impulse, Focusai, and Foundries.io. Additionally, Qualcomm announces the completion of the Augentix acquisition, primarily aimed at strengthening its camera and vision segment.

Toolkit for Developers: Arduino meets Edge Impulse and Foundries.io

At the core is a promise to developers: less friction between prototyping, model training, and secure operation in the field. Qualcomm describes a unified software architecture covering Linux, Windows, and Android, enabling the same basic building blocks to be used across various device classes.

Gallery

Arduino is intended to keep the entry simple, Edge Impulse to support training and model management, and Foundries.io to pave the way to secure operation. Qualcomm is thus targeting not only major customers but explicitly also smaller OEMs and independent developers who want to bring devices to series production without large platform teams.

Two New Q Processors: Multi-Camera and Multimedia Come Into Focus

With the new Dragonwing processors Q-8750 and Q-7790, Qualcomm focuses on on-device AI in devices that today often still rely heavily on cloud services. The Q-8750, with an AI performance of 77 TOPS (INT4/8/16 and FP16), is also designed to locally run larger models. Qualcomm cites large-language models with up to 11 billion parameters as the target size. A camera architecture for up to twelve physical cameras and three 48-megapixel ISPs is aimed at drones and multi-angle vision systems.

The Q-7790 is designed as a smaller variant for "everyday devices," such as smart cameras, AI TVs, media hubs, and video conferencing systems. Qualcomm specifies the on-device AI performance at 24 TOPS. It also features multimedia capabilities such as two 4K60 displays, 4K60 encoding, and 4K120 decoding, including AV1 hardware decoding. For security-critical environments, Qualcomm cites Secure Boot, a Trusted Execution Environment, and a management engine as protection mechanisms.

The new Q processors clearly target device classes with a strong focus on camera and multimedia, such as drones, video, and collaboration systems. The Dragonwing-IQ family, on the other hand, addresses typical industrial edge systems like industrial PCs, HMIs, controllers, and edge controllers.

Augentix and the Insight Platform: Cameras as a Lever for Edge Video

With the completion of the Augentix acquisition, Qualcomm is expanding its position in the field of image processing. According to Qualcomm, Augentix provides image processors and SoCs for IP cameras and networked video systems. This is expected to enable the company to deliver more customized chips for camera designs and broaden its own roadmaps for smart cameras.

In parallel, Qualcomm is bringing the Insight Platform to the forefront. Qualcomm describes it as a service for video intelligence that combines edge AI with a dialog-oriented, LLM-based query. The goal: to modernize existing installations, for example, via edge AI boxes or AI-capable cameras, and to make video data usable in real-time for security and operational processes. (mc)

Subscribe to the newsletter now

Don't Miss out on Our Best Content

By clicking on „Subscribe to Newsletter“ I agree to the processing and use of my data according to the consent form (please expand for details) and accept the Terms of Use. For more information, please see our Privacy Policy. The consent declaration relates, among other things, to the sending of editorial newsletters by email and to data matching for marketing purposes with selected advertising partners (e.g., LinkedIn, Google, Meta)

Unfold for details of your consent