By 2026, edge AI will no longer be a concept that needs to be "defined," but a reality that is being "quantified"—and the speed and scale of this quantification far exceed the expectations of most analysts a year ago. Edge AI is moving from "proof of concept" to "large-scale deployment."
I. Qualcomm's Strategic Move: Dragonwing Dual-Processor Chip Released to Boost Edge AI
On the eve of the 2026 IFA Berlin, Qualcomm officially launched two processors , the Dragonwing Q-2390 and IQ-2390. The Q-2390 highly integrates edge AI inference, cellular connectivity, positioning, and display support into a single chip, aiming to lower the development threshold for edge AI devices. The IQ-2390 focuses on enhancing machine vision acceleration, TSN (Time-Sensitive Networking) support, and has a wide operating temperature range of -30°C to +115°C , designed for demanding scenarios such as industrial automation, machine vision, and energy management.
Qualcomm's core intention in this announcement is to reduce the complexity of implementing edge AI in industrial scenarios through chip-level integration and industrial hardening.
II . Edge AI Market: A Trillion-Dollar Track Accelerating
The market size of edge AI is expanding at an astonishing rate.
According to a recent report by Global Market Insights, the global edge AI market is valued at approximately $25.2 billion in 2025, projected to reach $30.9 billion in 2026, and is expected to grow to $225.5 billion by 2035, representing a compound annual growth rate (CAGR) of 24.7% from 2026 to 2035. Data from Mordor Intelligence also corroborates this trend: the edge AI hardware market is projected to grow from $25.08 billion in 2025 to $30.74 billion in 2026, reaching $68.73 billion in 2031 .
The global edge AI market (hardware + software + services) is approaching $47-50 billion in size by 2026, with a year-on-year growth rate exceeding 28% . IDC predicts that the overall edge computing market will reach $450 billion by 2029, with AI being the primary growth engine .
From a regional perspective, the Asia-Pacific region is the fastest-growing edge AI market globally . China, as a core growth engine, leads the world in shipments of edge devices equipped with domestically produced computing chips. Bulk procurement in areas such as smart security, smart cities, and factory automation continues to drive demand, with the domestic market growing at an annual rate exceeding 34% .
III . The AI-driven transformation of the module industry: From "connectivity" to "computing + connectivity"
The explosive growth of the edge AI market is profoundly reshaping the competitive logic of the module industry.
Traditional module manufacturers heavily rely on the number of connections and shipment growth, but this model is encountering bottlenecks. In the first quarter of 2026, shipments of AI embedded cellular IoT modules declined by 17% year-on-year, marking the first decline after 12 consecutive quarters of growth in the market. The penetration rate of AI modules in global cellular IoT modules is currently only about 6%—meaning that edge AI modules are still on the verge of explosive growth, rather than a red ocean market .
Meanwhile, leading companies in the industry are accelerating their transformation towards "connectivity + intelligence." Quectel Wireless Solutions' revenue in the first half of 2026 reached RMB 15.259 billion, a year-on-year increase of 32.16%, with automotive, intelligent, and solutions businesses accounting for 46.76% of revenue. Furthermore, the gross profit margin of this segment (21.42%) is significantly higher than that of traditional communication module businesses (16.28%) . Andrew Zignani, Senior Research Director at ABI Research, pointed out that wireless connectivity is no longer just about connecting devices; it's also about enabling scalable edge AI in consumer, enterprise, and industrial markets .
IV . Trends in Module Products Driven by Edge AI
The large-scale deployment of edge AI is reshaping the definition of module products from three dimensions.
Trend 1: From "Communication Modules" to "Computing Modules" . The integration of AI edge computing chips has become the core differentiator for modules in 2026. Shipments of modules supporting edge inference exceeded 80 million units in 2025, and this proportion is expected to rise to 35% of total module shipments by 2030. The compound annual growth rate of high-performance computing modules is projected to reach 60% over seven years, and the proportion of intelligent and AI modules in cellular IoT modules will continue to increase.
Trend Two: Industrial-grade wide operating temperature range becomes the "entry ticket ." Wide operating temperature capability reflects the stringent reliability requirements of modules in industrial edge AI scenarios. Industrial-grade modules typically require an operating temperature range of -40℃ to +85℃. In scenarios such as oil and gas, power, and outdoor infrastructure, wide operating temperature capability directly determines whether the equipment can operate stably in real-world environments.
Trend 3: Multi-technology integration becomes standard . Edge AI scenarios often require simultaneous support for multiple connectivity methods—industrial automation needs Wi-Fi/Bluetooth for inter-device communication, energy management needs PLCs to solve the problems of transmission through walls and over long distances, and industrial gateways need cellular networks to upload data to the cloud. Modules with single communication methods are being replaced by integrated solutions that combine multiple modes .
V. Qogrisys's Product Strategy: AI Capabilities Connected to the Base
Amid the industry wave of module manufacturers moving towards AI, QOGRISYS's strategic path clearly demonstrates the industry logic of integrating "connectivity + computing".
Qogrisys's strategic path can be summarized as follows: using high-speed connectivity as the foundation and edge AI as the incremental factor, to build an intelligent module product matrix covering multiple scenarios.
Wi-Fi 7 Module: Full Product Lineup
In the Wi-Fi 7 field, O2072PM and O2072PB, two Wi-Fi 7 + Bluetooth 6.0 combo modules based on the Qualcomm QCC2072 chip, fully support 4K QAM, 320MHz channels, MLO (Multi-Link Operation), with a peak rate of 5.8Gbps and eMLSR enhanced multi-link switching. The O2072PM uses an M.2 Key E interface (22×30mm), making it suitable for high-end robots and industrial equipment. From an industry trend perspective, Wi-Fi 7 is rapidly entering a large-scale deployment phase. IDC data shows that in the first quarter of 2026, Wi-Fi 7 already accounted for 44.5% of global enterprise WLAN-dependent AP revenue. The growth of Wi-Fi 7 is no longer limited to routers and enterprise APs; IoT devices are becoming an important source of growth in the next stage .
Multi-platform and multi-scenario coverage
Qogrisys's core R&D team has long been deeply involved in mainstream global wireless chip platforms such as Qualcomm, Realtek, and MediaTek, accumulating full-stack experience from chip bring-up and RF calibration to mass production testing . This multi-platform, full-stack technical capability enables Qogrisys to provide compatible connectivity solutions across different customers' main control solution ecosystems.
In terms of application scenarios, Qogrisys's product matrix has extended to multiple core areas of edge AI:
industry and intelligent manufacturing , modules such as O9201PM have completed driver adaptation and full verification on domestic platforms such as RK3588, Allwinner, and Rockchip, and can be widely used in scenarios such as industrial tablets, edge computing gateways, industrial control equipment, smart retail terminals, and digital signage.
robotics and embodied intelligence , the ultra-high bandwidth and ultra-low latency provided by Wi-Fi 7 form an invisible infrastructure for cloud-edge-device collaboration—robots upload sensor data to edge servers in real time for VLA model inference, receive decision commands, and execute them instantly. Qogrisys's Wi-Fi 7 module product line already covers the needs of high-end robots and industrial equipment.
VI. Trend Analysis: Three Certain Directions for Edge AI Modules
Based on Qualcomm's chip release trends, three definite trends can be identified in the field of edge AI modules:
First, AI capabilities will become a standard feature rather than an optional feature for modules. As the report "Edge Intelligence 2026: The First Year of Large-Scale Deployment" points out, 2026 has become a crucial turning point for edge intelligence, moving from proof-of-concept to large-scale deployment. The market space for pure connectivity modules lacking AI acceleration capabilities will continue to shrink. The seven-year compound annual growth rate of intelligent modules and AI modules reached 60% and 28% respectively , and this difference in growth rate is the strongest evidence of this.
Second, industrial scenarios represent the most certain high-value market for edge AI modules. The industrial AI accelerator chip market is expanding at an average annual growth rate of 40%. Industrial visual inspection is migrating from traditional algorithms to deep learning, and predictive maintenance is generating explosive demand for edge inference computing power.
Third, the convergence of "connectivity + computing" modules will become the infrastructure of the AIoT era. Wi-Fi 7 provides a connection base with ultra-low latency and ultra-high throughput, while edge AI provides localized intelligent decision-making capabilities. When the coverage density of Wi-Fi 7 and the computing power density of edge AI converge at the device end, the module will be upgraded from a "communication component" to an "intelligent computing node".
