Analyst(s): Olivier Blanchard
Publication Date: September 15, 2026
What Is Covered in This Article:
- Qualcomm introduced the Dragonwing Q-2390 and IQ-2390 processors on September 1, 2026, then used IFA Berlin (September 4-8) for on-site demonstrations — its most visible push yet to put edge AI silicon into everyday consumer and industrial devices rather than smartphones.
- The Q-2390 targets consumer/commercial AIoT (retail point-of-sale, smart appliances, home robots, fitness equipment, access control), while the IQ-2390 opens Qualcomm’s new “IQ2” industrial series for machine vision, building management, and factory automation, with an extended -30°C to +115°C operating range and support through 2036.
- Four partners — Eight Development, Fibocom, SECO, and Engicam — announced modules, SoMs, and reference designs built on the new chips, with evaluation kits expected in Q1 2027.
- Fortune Business Insights projects the edge AI semiconductor market Qualcomm is chasing here at $29.85 billion in 2026, growing to $107.86 billion by 2034 (17.4% CAGR) — context for why Qualcomm keeps expanding Dragonwing even as its Hexagon NPU here (up to 1.1 TOPS) sits well below rival parts like MediaTek’s Genio 420 (7.2 TOPS) or Nvidia’s Jetson Orin NX (100 TOPS).
The News: Qualcomm introduced new Dragonwing Q-2390 and IQ-2390 processors on September 1, 2026, ahead of IFA Berlin. Both chips share a quad-core Kryo CPU built on Cortex-A78/A55 cores (up to 1.9GHz), an Adreno 704 GPU, a real-time RISC-V MCU (SiFive E61), dual-channel LPDDR4x memory, and support for Android, Linux (Yocto/Ubuntu), and Zephyr RTOS. AI-enabled “AB” variants add a Hexagon NPU rated up to 1.1 TOPS.
The two parts are aimed at different buyers: The Q-2390 is pitched at consumer and commercial AIoT devices such as retail point-of-sale systems and kiosks, access control, smart appliances, agricultural equipment, home robots, fitness equipment, and enterprise terminals. It features on-device AI, camera, and display support, optional LTE Cat 4 and GPS, and both wired and wireless connectivity. The IQ-2390, on the other hand, is the first chip in a new “IQ2” series built for industrial edge AI. It features dual Gigabit Ethernet with Time-Sensitive Networking, a more robust -30°C to +115°C operating range, shock and vibration durability, and a commitment to keep both parts available through 2036 — a longevity guarantee that matters more to industrial customers designing decade-long deployments than to consumer device makers.
Four hardware partners announced products built on the new silicon alongside Qualcomm’s launch: Eight Development is building general-purpose designs on the Q-2390; Fibocom is developing its SC236 smart module, aimed at payment terminals and industrial handhelds, on the same chip. On the industrial side, SECO is pairing the IQ-2390 with its Clea software framework across a Compact Vision 5 machine-vision board and a single-board computer, while Engicam is building a System-on-Module version.
Jeff Arnold, Qualcomm’s VP & GM for Auto Telematics, Industrial & Embedded IoT, Consumer and Connectivity, framed the launch as addressing categories that “still face barriers around cost, complexity and integration,” with the goal of helping “intelligence to scale across a much broader range of devices, from retail systems, smart appliances and consumer robots to industrial controllers, machine vision systems and critical infrastructure.” Evaluation kits are expected in Q1 2027.
The launch landed in a week when other Snapdragon silicon was also visible across IFA in partner products: Motorola’s Watch Ultra powered by the Snapdragon W5+ Gen 1, Honor’s robot-phone concept powered by the Snapdragon 8 Elite Gen 5, and Lenovo’s IdeaPad Vibe, which offers a Snapdragon X option. Together with the Dragonwing announcements, IFA again served as a showcase for Qualcomm silicon diversity and presence across critical device segments.
Separately, and unrelated to IFA, Qualcomm also announced a multi-generation custom-silicon and optical-connectivity partnership with AWS for AI data center infrastructure on September 8, which deserves its own report.
What Qualcomm’s Focus at IFA Berlin Signals About the Company’s Intelligent Edge Strategy
Analyst Take—Why IFA Was the Right Stage for This Launch: Dragonwing chip launches usually surface at CES or embedded-systems trade shows aimed at engineers. Putting Q-2390/IQ-2390 demos on the IFA floor this year signals that the Q-2390’s target list (smart appliances, home robots, retail kiosks, fitness equipment) increasingly reads like an IFA exhibitor directory: For starters, Qualcomm didn’t just focus on the developer audience that Dragonwing announcements more typically target. Second, the company went straight to the consumer-electronics OEMs who fill that hall, and in plain view of the buyers (appliance makers, robot startups, kiosk vendors) who decide what chip goes into next year’s products. At this point in the Dragonwing story, that seems like a smart approach.
The Diversification Math Behind the Chip
At its 2026 Investor Day, Qualcomm targeted more than $14 billion in IoT revenue by fiscal 2029, split roughly $8 billion for industrial, networking, and robotics, and $6 billion for personal AI and compute. For reference, this is part of a broader push toward $40 billion in non-handset QCT revenue.
IoT revenue grew to $1.83 billion in fiscal Q3 2026 alone, still some way from the 2029 target, but already up 9% year over year, and with an industrial design-win pipeline that may already exceed $7 billion. Now, two chips and four launch partners won’t move that pipeline by much on their own, but Dragonwing’s consumer/industrial split maps directly onto the two IoT sub-targets Qualcomm has already committed to publicly, which makes this look less like an isolated product refresh and more like restocking a shelf Qualcomm has already told Wall Street it needs to fill. Additionally, momentum reads like validation, and validation tends to fuel momentum, so Qualcomm’s successful splash at IFA is also designed to convey a very specific message to the market, and in that regard, I think that we can add this exercise to the wins column.
The Spec Qualcomm Would Rather You Not Compare Directly
I need to address something that keeps coming up in discussions about the launch, though: The Hexagon NPU in AI-enabled variants tops out at only 1.1 TOPS, which seems like a modest number next to NXP’s i.MX 8M Plus (2.3 TOPS), MediaTek’s Genio 420 (7.2 TOPS), or, in a different weight class entirely, Nvidia’s Jetson Orin NX (100 TOPS). But that comparison isn’t fully fair: For starters, Jetson Orin NX serves compute-heavy robotics and vision workloads at a different price and power point than retail kiosk or smart-appliance chip use cases. While TOPS-focused comparisons often make sense in PC and handset applications, the AIoT is a completely different segment with a far broader range of form factors, use cases, and spec requirements. Unfortunately, Qualcomm’s own framing (“advanced edge AI within reach of more customers”) invites exactly this kind of apples-to-oranges scrutiny, which can lead reviewers and competitors to draw comparisons that shouldn’t be made.
The more defensible pitch here isn’t raw AI throughput, it’s the combination of price tier, software portability across Android/Linux/Zephyr, and — for the IQ-2390 specifically — a 2036 longevity commitment that industrial customers designing decade-long deployments will weigh more heavily than an out-of-context TOPS figure.
Market Sizing Gives This Some Real Runway
The edge AI semiconductor market is expected to grow to exceed $100 billion by 2034, at a roughly 17-20% CAGR. That’s a market-wide figure, not a Qualcomm-specific forecast, and it says nothing about what share Qualcomm can realistically capture against NXP, MediaTek, Renesas, and Nvidia, all of which compete in some slice of this same space. But the tailwind is there, Qualcomm’s IP looks solid, and so I see a credible path to growth for Qualcomm in that space. Painted against this broader tapestry and timeframe, Qualcomm’s Dragonwing announcements at IFA feel like bedrock foundations for what comes next, and the key here is that even a thin share of a market nearly quadrupling by 2034 is a more durable growth line than overreliance on flagship-phone silicon. This insight isn’t new (it has been driving Qualcomm’s diversification strategy for the better part of the last decade now), but Qualcomm’s focus at IFA feels like yet another milestone in the company’s expansion into fresh high-value intelligent edge segments.
What to Watch:
- Whether Eight Development, Fibocom, SECO, and Engicam ship commercial Q-2390/IQ-2390 modules and reference designs on schedule once Q1 2027 evaluation kits go out, and whether additional OEMs sign on before then.
- Whether Dragonwing design wins show up in Qualcomm’s IoT segment revenue in the fiscal quarters ahead, against the >$14 billion by fiscal 2029 target (industrial/networking/robotics plus personal AI/compute), Qualcomm laid out at its 2026 Investor Day.
- Whether NXP, MediaTek, Renesas, or Nvidia respond with comparably priced, similarly long-lived parts at the low-TOPS end of edge AI, sharpening competition in the tier Dragonwing just expanded into.
- Whether Qualcomm’s September 8 AWS data center silicon partnership — announced the same week but unrelated to IFA — ends up as the bigger diversification story of the two, meriting separate Futurum coverage.
For more information, see the press release on Qualcomm’s newsroom.
Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of Futurum as a whole.
Other Insights From Futurum:
Qualcomm’s Investor Day 2026: Agentic and AI Inference To Drive 2x Revenue Growth by 2030
Qualcomm Q3 FY 2026: Automotive Growth Offsets Handset Weakness
Qualcomm Unveils Future of Intelligence at CES 2026, Pushes the Boundaries of On-Device AI
Author Information
Olivier Blanchard is Research Director, Intelligent Devices. He covers edge semiconductors and intelligent AI-capable devices for Futurum. In addition to having co-authored several books about digital transformation and AI with Futurum Group CEO Daniel Newman, Blanchard brings considerable experience demystifying new and emerging technologies, advising clients on how best to future-proof their organizations, and helping maximize the positive impacts of technology disruption while mitigating their potentially negative effects. Follow his extended analysis on X and LinkedIn.
