Quality WiFi7 Module & WiFi HaLow Module manufacturer

On November 3, 2026, Universal Scientific Industrial (USI) officially announced its Wi-Fi 8 product roadmap, expecting to launch Wi-Fi 8 modules based on SiP packaging in the first quarter of 2027. This news sparked considerable discussion within the module industry , and we've also been following this development . We'd like to discuss a more practical topic: Wi-Fi 8 is the "next step" worth paying attention to, while Wi-Fi 7 is currently the "mainstay" being delivered in large quantities .   For hardware teams that are defining their products, recognizing this time lag is more important than chasing trends.   I. The Big Moves of Wi-Fi 8: How Far Away is Mass Production? Universal Scientific Industrial (USI)'s roadmap reveals two key pieces of information. First, a clear timeline: products will launch in Q1 2027. Second, a shift in technological focus: Wi-Fi 8 (IEEE 802.11bn) will no longer prioritize peak speeds as its primary selling point, but will instead focus on ultra-high reliability (UHR), energy efficiency, and a consistent user experience. Key technologies listed in the announcement include Co-BF/Co-SR, Dynamic Subchannel Operation (DSO), and physical layer technologies such as Distributed Resource Unit (DRU), Enhanced Long-Range Transmission (ELR), and Unequal Modulation (UEQM). These technologies target specific scenarios: miniaturized terminals with extremely limited antenna space, limited battery capacity, and highly complex radio frequency environments, such as AI glasses, smart rings, and TWS earphones. However, as a module solution provider, our habit when looking at this kind of news is to follow the timeline and see where the standard is finalized and the certification is opened. According to ABI Research's tracking report, the standardization progress of 802.11bn is as follows: Draft 2.0 entered the working group vote in August 2026, the Draft 3.0 letter ballot is scheduled for January 2027 , the Draft 4.0/initial standards association ballot is in May 2027 , final working group approval is expected in March 2028 , and IEEE Standards Board (RevCom) ratification is in May 2028. The Wi-Fi Alliance's Wi-Fi 8 certification project is not expected to start until January 2028 . What does this mean? If a product launches in Q1 2027, it will be facing a stage where standards are not yet finalized and certification channels are not yet open. Draft 3.0 has not yet been voted on, Draft 4.0 has not yet entered the SA ballot, and the Wi-Fi Alliance's certification program has not even set its start date for another year. For the vast majority of terminal products that need to go through type approval and alliance certification before they can be launched, this timeframe does not meet the conditions for mass production. Dell'Oro Group shares this view: Wi-Fi 7's market share exceeded 50% by September 2026, while early revenue from Wi-Fi 8 is not expected until 2027 at the earliest. ABI Research directly used the phrase "lucrative opportunities and unprecedented challenges" in the title of its report, indicating that opportunities and challenges coexist, but the window of opportunity is in 2028, not in Q1 of 2027. II. Why Wi-Fi 7 is the realistic choice for customers today Looking only at news trending topics can easily create a misconception. But the data presents a completely different picture. According to Dell'Oro Group data, by September 2026, Wi-Fi 7 devices accounted for over 50% of the global WLAN market share . In the Chinese router market, Sandalwood's e-commerce monitoring data provides a more concrete picture: by September 2026, Wi-Fi 7's market share rose from 26% to 32% , while Wi-Fi 5's fell from 22% to 12%, indicating a rapid shift in market structure towards next-generation protocols. The BE3600 specification has become the mainstream adoption level for Wi-Fi 7, with its share rising from 13% to 17%. The replacement cycle in the home consumer market has entered an accelerated phase, rather than just beginning. At its earnings call in early 2026, Realtek gave its forecast for the chip sector: Wi-Fi 7 penetration in PCs and routers is expected to double in 2026, reaching 30% in PCs and 25% in routers. Considering that Realtek is one of the major suppliers of Wi-Fi chips, this forecast reflects the capacity allocation rhythm of the upstream supply side. For customers, a more direct criterion is this: if your current product project is planned for mass production between the end of 2026 and 2027, Wi-Fi 7 is the only option with a complete supply chain support. Wi-Fi 8's modular solution won't be ready for evaluation until certification opens in 2028 at the earliest. During this one-year window, customers' products cannot wait, and production lines cannot stop. III. OFEIXIN Wi-Fi 7 Strategic Deployment Returning to our own products , our two Wi-Fi 7 modules, based on the Qualcomm QCC2072 platform, cater to the needs of two mainstream hardware architectures currently on the market. Our initial motivation for this dual-package layout was simple: different customer decision-making stages necessitate different module forms. The O2072PM uses an M.2 2230 Key E slot and is designed for devices with standard M.2 interfaces, such as gateways, routers, and industrial PCs. These customers typically require pluggable modules for easy maintenance and upgrades, and have specific requirements for driver compatibility with the Intel x86 platform. The O2072PM is already compatible with kernels 5.15.24+, 6.1.99, and 6.12.0, and also supports OpenWrt version V1. Customers can immediately debug the module without having to figure out the driver from scratch , which is where we invested the most in software adaptation. For many customers evaluating Wi-Fi 7 modules, the first hurdle isn't RF performance, but whether the driver will work and the SDK will be compatible. We've addressed this issue beforehand. The O2072PB is a 13×15×2.3mm surface-mount package, designed for PCB space-constrained scenarios such as FTTR gateways, sub-routers, and high-density IoT terminals. It supports Wi-Fi via PCIe and Bluetooth via UART/PCM, integrating Bluetooth 6.0 and supporting multiple operating modes including AP/AP+STA/P2P/Monitor. Surface-mount customers typically lock in the module size and pin definitions during the onboard design phase, giving them a longer selection window than slot-based designs . This is why the O2072PB needs to coexist with the O2072PM. If a customer locks in a surface-mount solution at the schematic stage and wants to switch to a slot-based solution before mass production, the redesign costs will be very high. Both modules share the same core Wi-Fi 7 capabilities: 802.11be, 2×2 MIMO, tri-band 2.4/5/6GHz, 320MHz bandwidth in the 6GHz band, 4096-QAM, and a peak speed of 5.8Gbps . The only differences lie in their physical form factor and interfaces; customers can choose according to their hardware architecture. It's worth mentioning the engineering value of eMLSR (Enhanced Multi-Link Single Radio). The O2072PM supports 320MHz eMLSR. In a single-radio hardware architecture, it monitors the status of multiple frequency band links via an auxiliary radio, quickly switching to a backup link when the primary link encounters interference, without needing to go through a complete scan-association process. For cost-sensitive and space-constrained designs like FTTR sub-routers, eMLSR provides a feasible path to achieve near-multi-radio latency performance on a single-radio architecture . This is also one of the most noteworthy engineering improvements of Wi-Fi 7 compared to Wi-Fi 6, changing the cost structure of high-density networking. IV. A variable to be aware of: the impact of memory shortage A recent report by Dell'Oro Group makes a noteworthy assessment: the shortage and price increase of DDR4 memory chips may accelerate the adoption of some enterprise-grade Wi-Fi 8 systems, as suppliers are incentivized to reduce their reliance on high-cost memory by switching designs. Realtek also explicitly mentioned in its earnings call that it faces the challenge of "memory shortages and price increases" in 2026. However, this does not contradict the statement that "Wi-Fi 7 remains the mainstay of shipments." Dell'Oro explicitly stated in the same report that Wi-Fi 7 revenue growth will continue, with peak annual revenue expected to exceed the total revenue of Wi-Fi 6E over its entire lifecycle . The memory shortage is driving some enterprise customers to assess the feasibility of Wi-Fi 8 as early as 2027, while the replacement cycle for Wi-Fi 7 in numerous scenarios such as consumer routers, industrial control systems, CPEs, and FTTRs is only just entering the accelerated phase shown by Sandalwood data. For customers, a more realistic assessment is: if your product is planned to enter mass production before the first half of 2027, Wi-Fi 7 is currently the only option that can guarantee delivery schedule. If the product cycle is longer, cost sensitivity is lower, and it targets high-end enterprise scenarios, you can pay attention to the standardization progress of Wi-Fi 8, but the module selection decision will still be made after the second half of 2027. In conclusion Universal Scientific Industrial's (USI) Wi-Fi 8 roadmap provides valuable forward-looking information. It tells the industry that the next-generation standard prioritizes reliability over speed, with module form factors evolving towards SiP miniaturization, and target applications expanding from mobile phones/PCs to AI glasses and wearable devices. Oufexin is also monitoring the evolution of Wi-Fi 8 and will promote modular adaptation as the standard is finalized and certification is opened. However, between the blueprint and mass production lies four stages: standard finalization, certification openness, chip maturity, and module adaptation . ABI Research's timeline shows that all four stages will be completed by May 2028. Meanwhile, Wi-Fi 7's shipment data reflects what is happening now: a global market share exceeding 50% by September 2026 and a 32% sales share in the Chinese router market . Behind these figures is a well-established supply chain and terminal projects being delivered in batches. For Oufexin, the key is that when customers need a Wi-Fi 7 module that can be immediately integrated into a project, has mature driver compatibility, and offers two packaging options, we already have one on our shelf . In the module industry, delivery speed is far more important than news pace. Data sources for this article: ABI Research, Dell'Oro Group, Sandalwood, and publicly available information from Realtek's earnings calls. Information on Oufexix products is from official specification documents.  

On October 27, 2026, the long-standing shortage of domestically produced Wi-Fi AP chips was broken. The launch of the WQ9301 from Wuqi provided a realistic foundation for the mass production and adoption of "all-domestic AP solutions" for the first time. For terminal manufacturers, however, the chip itself is only the starting point—what truly determines whether a product can be launched on time and shipped stably is the engineering capability of the module stage . I. Why is the AP chip the most difficult part of domestic Wi-Fi development? Wi-Fi chips are functionally categorized into STA (Station) and AP (Access Point). Over the past few years, breakthroughs in domestically produced Wi-Fi have primarily focused on the STA field—Wi-Fi modules in set-top boxes, cameras, and IoT devices—where domestically produced chips have already captured a significant market share. However, the situation is quite different for AP chips. The AP chip is the central hub of a Wi-Fi network, handling concurrent access from multiple devices, scheduling channel resources, managing QoS queues, and ensuring overall network stability. It is widely acknowledged in the industry that the R&D difficulty and investment required for AP chips are far greater than those for STA chips. According to data from Shangpu Consulting, Broadcom, Qualcomm, and MediaTek collectively hold over 70% of the global high-end Wi-Fi chip market share, while domestic companies account for only about 4.86%. In core AP scenarios such as routers and smart gateways, the penetration rate of domestically produced chips has long remained low. The direct consequence of this situation is that domestic manufacturers of routers, gateways, and optical modems have almost no domestically produced chips to choose from in their AP product lines. Supply chain security is one layer of pressure, and cost control is another—without alternatives, bargaining power is extremely limited. The significance of the WQ9301 lies in the fact that it has, for the first time, transformed the concept of "domestic AP chip" into a product that can be inquired about, sampled, and evaluated for mass production. II. To what extent has the Wuqi WQ9301 achieved this? The WQ9301, launched by Wuqi Microelectronics, is one of the very few mass-produced RISC-V architecture Wi-Fi 6 router chips in China. Based on its publicly available specifications, it is positioned to directly compete with mainstream Wi-Fi 6 access point (AP) solutions.   The architecture employs a RISC-V multi-core instruction set, runs the IEEE 802.11ax protocol stack, and is backward compatible with 802.11 a/b/g/n/ac. In terms of speed, it supports 160MHz bandwidth and 1024QAM modulation, utilizes DL/UL MU-MIMO and OFDMA technologies, and achieves a maximum dual-band concurrent speed of 3Gbps at the physical layer, targeting 4K/8K video and large-scale online gaming scenarios. The radio frequency (RF) is the chip's core differentiator: It features a self-developed 2.4GHz and 5GHz dual-band RF circuit and a unique high-performance CMOS PA technology, controlling power consumption and heat generation while ensuring transmit power. In terms of integration, a single chip integrates baseband, RF, and FEM, while also supporting external FEM solutions to meet diverse needs. More importantly, there's the progress of industrialization. The WQ9301 has passed authoritative industry certification and domestic production assessment, and has already achieved small-batch supply to ZTE Microelectronics and Changhong. It is expected to be adopted by major terminal customers such as China Mobile by 2026. Simultaneously, Wuqi is advancing the development of Wi-Fi 7 AP chips, aiming to complete Wi-Fi Alliance certification as early as the end of 2026. This means that the WQ9301 is not just a chip that remains in the release draft, but has already entered the evaluation and adoption process by real customers. For terminal manufacturers that make routers, smart gateways, and set-top boxes, there is finally a domestic AP solution that they can seriously evaluate. III. From Chips to Modules: Ofeixix Role is "Translation" A Wi-Fi chip is shipped as a packaged bare die, but terminal manufacturers need a module that can be plugged in and run immediately. Antenna matching, RF calibration, driver adaptation, heat dissipation design, protocol stack verification, certification testing—if these tasks are not completed, the chip's performance cannot be released, and the terminal product cannot be launched . This is what OufiXin does. It's a proven fact that the Wuqi WQ9201 chip has been adopted by multiple China Mobile set-top boxes and has achieved mass production. The O9201SB module developed by Oufexin based on this chip is a crucial link in this mass production chain. This module, measuring only 13×15mm, supports 2×2 MU-MIMO and dual-band concurrency, addressing the pain points of video stuttering and game latency in set-top box scenarios. From chip to module to bulk procurement by operators, this path has been successfully established. Designed for industrial applications, the O9201PM module from Oufexin adopts an M.2 interface and PCIe protocol, and has completed driver adaptation on domestic platforms such as RK3588, Allwinner, and Rockchip. Power consumption is reduced by 40% when Wi-Fi and Bluetooth work together. This demonstrates that Oufexin's capabilities extend beyond consumer-grade modules, and it also has delivery experience in domestic IT innovation and industrial control scenarios. When the WQ9301 entered the modularization stage, Oufexin's existing mass production experience with the Wuqi solution constituted a real first-mover advantage. New module manufacturers would need to go through the entire process of RF calibration, driver adaptation, and certification testing from scratch, while Oufexin had already accumulated a large amount of reusable engineering data and production line experience on the Wuqi platform. IV. Market Window: Wi-Fi 6 Remains the Core Market, and the Time is Ripe for Domestic APs. A common misconception is that Wi-Fi 7 is rapidly becoming widespread, so is it too late to develop Wi-Fi 6 AP chips now? The data tells us otherwise. According to a report by Luotu Technology, in the first half of 2026, Wi-Fi 6 accounted for 57.5% of the online market share for home routers in China, while Wi-Fi 7 accounted for 30.1%. Wi-Fi 6, with its cost-effectiveness advantage in the core price range of 100-300 yuan, covers the replacement needs of the vast majority of ordinary families and remains the absolute market mainstream. More macro-level data comes from Shangpu Consulting: in 2024, China's Wi-Fi chip shipments exceeded 2 billion units, with a market size approaching 40 billion yuan; it is projected that the Chinese market size will reach 100 billion yuan by 2030, but domestic companies currently only hold about 4.86% of the high-end chip market share. This figure itself illustrates the substitution potential of domestic AP chips—not a competition for existing market share, but a climb from almost zero to a double-digit market share. For Oufexin's target customers—set-top box manufacturers, gateway manufacturers, and industrial control equipment manufacturers—Wi-Fi 6 is currently the product line with the largest shipment volume, and the availability of domestic AP chips is directly related to BOM cost and supply chain security. V. Conclusion: When choosing chips, it's even more important to choose module partners who have already proven their capabilities. While the options for domestically produced Wi-Fi AP chips are indeed increasing, more options do not equate to easier implementation. The transition from "sampleable" to "mass-producible module" for a new chip involves a series of significant challenges, including RF debugging, driver adaptation, certification cycles, and production line yield ramp-up. If a terminal manufacturer chooses a module supplier without experience in mass-producing that chip platform, the project risks will not automatically disappear simply because "domestic chips are available." Oufeixin's core value lies not in its "distribution" of Wuqi's chips, but in its completion of a full engineering loop from chip to module on the Wuqi platform, which has been validated by mass shipments of China Mobile set-top boxes. When the WQ9301 enters the modularization stage, this proven engineering system can be directly reused, significantly reducing the adoption cycle for end customers. Congratulations again to Wuqi on the successful deployment of its WQ9301. For router, gateway, and set-top box manufacturers evaluating domestic AP solutions, the most pragmatic choice is not to wait for "more chip options," but to find a module partner that has already successfully run mass production using domestic chips . The combination of Oufexin and Wuqi provides a starting point for immediate evaluation. Data source notes : Wuqi WQ9301 parameters and industrialization progress are from reports by China Electronics Network and IT Times, as well as Wuqi's IPO prospectus; Oufexin module and mass production information are from Oufexin's official public information and reports by Electronic Engineering Times; market size data is from Shangpu Consulting's "WiFi Chip Industry Research Report (2026)"; router market structure data is from Luotu Technology's online market statistics for the first half of 2026.  

In October 2026, the "Opinions on Developing New Productive Forces" issued by the Central Committee of the Communist Party of China and the State Council listed humanoid robots as one of the key tasks for the application of next-generation intelligent terminals. Dai Jianjun, Deputy Director of the Innovation and Development Research Department of the Development Research Center of the State Council, pointed out in an interview with the China Economic Times that humanoid robots "are not isolated consumer products, but rather super carriers integrating hardware and software and spanning the entire industrial chain," capable of driving upstream computing power, chips, algorithms, and sensors, and simultaneously promoting the systematic upgrading of traditional supporting industries and manufacturing chains.   The industrial implications of this assessment are that the large-scale deployment of humanoid robots tests not only the manufacturing capabilities of the robots themselves, but also the maturity of the entire supporting supply chain. Dai Jianjun also pointed out the current core pain points: poor adaptability to unstructured environments, high costs of modifying the entire robot and its on-site components, and a high degree of scenario customization, making rapid deployment difficult . At the wireless connectivity level, these pain points directly translate into engineering requirements for link determinism, anti-interference capabilities, and integration efficiency.   1. In Dai Jianjun's judgment, which aspects are directly related to wireless connectivity? In Dai Jianjun's interview, there are three statements that have a technical mapping relationship with the Wi-Fi industry. Firstly, the implementation path is "starting with structured industrial scenarios and gradually expanding to semi-structured and home scenarios ." In structured factory environments, the dense metal structures and complex electromagnetic reflections make traditional Wi-Fi prone to jitter and momentary disconnections on a single frequency band. The primary requirement for robots in wireless links is not peak speed, but continuity of connection in interference-prone environments . Secondly, "poor adaptability to unstructured environments" is listed as a core pain point . Unstructured environments mean that robots need to perceive and make decisions in real time in dynamic spaces. Their bandwidth requirements for uplink data (vision, point cloud, force) and their sensitivity to the latency of control commands are significantly higher than those of industrial equipment with fixed workstations. Third, humanoid robots will "drive the development of intelligent manufacturing equipment in the short term ." As robots move from demonstration to mass deployment, the communication modules and radio frequency devices that support them will no longer be small-batch customized products, but will need to meet the requirements of large-scale industrial products with consistency, reliability, and supply stability. Key findings: The “poor adaptability to unstructured environments” and “systematic upgrade of manufacturing links” pointed out by Dai Jianjun correspond to two technical issues at the wireless connectivity level: link determinism in complex electromagnetic environments and quality convergence of communication modules from customized parts to mass-produced industrial parts. II. Which capabilities of Wi-Fi 7 substantially correspond to the above requirements? Among the core features of Wi-Fi 7, three are directly relevant to humanoid robot scenarios: Multi-Link Operation (MLO), 6GHz band support, and higher-order modulation and wider channels. However, it is necessary to distinguish between "technically corresponding" and "engineerably feasible." The industrial value of MLO lies in "uninterrupted connectivity," not "faster speed." In manufacturing environments with densely packed metal racks, AGVs and robots frequently experience jitter and momentary disconnections on a single frequency band. MLO allows devices to maintain two links simultaneously, switching traffic to the other when one band is interfered with. However, it must be objectively noted that MLO exists in two forms: STR (Simultaneous Transmit and Receive) and NSTR (Near-Switchover). As of 2026, most enterprise-level clients only support NSTR, and the proportion of industrial IoT devices supporting MLO remains limited. The industrial value of MLO is real, but the maturity of the client ecosystem will take time. The significance of the 6GHz band lies in spectrum isolation, but its availability in China is constrained by policy. The 320MHz channel relies on the 6GHz band, but China has not yet allocated the 6GHz band for Wi-Fi. This means that in domestic industrial scenarios, the peak bandwidth advantage of Wi-Fi 7 cannot be fully realized; what can actually be relied upon is the 160MHz channel and MLO redundancy capability under the 5GHz band. For module products targeting the global market, this regional difference needs to be clearly distinguished in selection and deployment strategies. A useful benchmark for implementation comes from the overall penetration rate of industrial Wi-Fi. Industry analysis indicates that by 2026, Wi-Fi 7 device shipments had surpassed Wi-Fi 6 shipments, with Huawei, H3C, and other manufacturers seeing industrial-grade Wi-Fi 7 AP shipments increase by over 120% year-on-year. In warehousing scenarios, Wi-Fi 7 upgrades the network from "best-effort" to industrial-grade determinism, with millisecond latency and extremely low packet loss significantly improving AGV scheduling efficiency. These data demonstrate that the industrial value of Wi-Fi 7 is being validated, but the validation scenarios are concentrated in warehouses, logistics centers, and manufacturing workshops with stable power supply and network infrastructure , rather than all industrial environments. Key findings: The core value of Wi-Fi 7 for humanoid robots lies in the link continuity provided by MLO (Mean Link Optimization), rather than peak speed. Given the constraint of the unopened 6GHz band in China, the actual competitiveness of module products depends more on the RF tuning capabilities of the 5GHz band and the engineering implementation quality of MLO. III. Specifications and Positioning of O2072PM/O2072PB from O2072PM O2Flytek's Wi-Fi 7 module, based on the Qualcomm QCC2072 (FastConnect C7700) platform, comes in two form factors: O2072PM uses the M.2 Key E standard interface (22×30mm), while O2072PB is a surface-mount package (13×15mm), designed for high-density onboard layouts. Core specifications include: 2T2R dual antennas, 2.4/5/6GHz tri-band, 320MHz channel bandwidth at 6GHz, peak data rate of 5.8Gbps, 4096-QAM modulation, and integrated Bluetooth 6.0. From the perspective of humanoid robot applications, the following three areas have value for industry analysis: The strategy of differentiating interface forms is engineeringly sound. M.2 Key E is a mature standard interface in embedded platforms, suitable for integration into robot main control boards with low design change costs; the surface-mount version O2072PB is geared towards space-constrained sub-modules or sensor nodes. Dai Jianjun pointed out that humanoid robots have a "high degree of scenario customization, making rapid promotion difficult," and the parallel use of the two packaging forms is actually using packaging flexibility to offset scenario fragmentation . The tri-band support and Bluetooth coexistence design are geared towards multi-wireless collaboration. The module explicitly supports Bluetooth and 2.4GHz WLAN sharing an antenna, while simultaneously operating in parallel with 5GHz/6GHz WLAN. In humanoid robots, Wi-Fi handles data transmission, while Bluetooth handles low-power sensor and remote control access; the quality of their coexistence directly impacts the overall wireless experience. This design choice aims for the collaborative operation of multiple wireless systems within a compact space , rather than maximizing the performance of a single link. The disclosure regarding platform compatibility is noteworthy. According to official information from Oufexin, the O2072PM is compatible with kernels 5.15.24+, 6.1.99, and 6.12.0 on the Intel x86 platform, but "further debugging and adaptation are needed" on the Arm platform . Customers should pay attention to whether the kernel version fully supports Wi-Fi 7. This disclosure highlights an industry reality: humanoid robots utilize diverse main control platforms (x86, Arm, and RISC-V are all under exploration), and the maturity of the Wi-Fi 7 module's drivers will directly affect its usability on different robot platforms. For robot main control solutions primarily based on the Arm architecture, this compatibility progress is a variable that must be considered during the selection process. IV. A necessary sober assessment: the pace of Wi-Fi 7 penetration in industrial scenarios When discussing the combination of Wi-Fi 7 and humanoid robots, there is a fact from the industry side that needs to be addressed. There is a significant discrepancy between Wi-Fi 7 shipment figures and actual deployments in industrial settings. Huawei cites Wi-Fi Alliance data stating that 625 million Wi-Fi 7 devices will be shipped globally in 2025, but the replacement cycle in industrial scenarios is much slower than in consumer electronics. The core considerations for industrial wireless selection are reliability and latency, while the market promotion of Wi-Fi 7 has long focused on "ultra-high throughput." This narrative mismatch may lead to slower-than-expected adoption of Wi-Fi 7 by industrial customers. Meanwhile, Wi-Fi 8 is expected to be released in 2028, and some industrial customers may choose to wait and see or jump directly from Wi-Fi 6 to Wi-Fi 8. The market size of humanoid robots also needs to be examined within reasonable expectations. Morgan Stanley raised its forecast for humanoid robot shipments in China to 50,000 units in June 2026, with a market size of approximately $2 billion; the forecast from China Business Industry Research Institute is more conservative, predicting shipments of approximately 38,000 units and a market size of 3.4 billion yuan in 2026. Regardless of which set of data is used, this size represents a meaningful segment for the Wi-Fi module industry, but it is not enough to change the overall penetration curve of Wi-Fi 7 in the short term . A more pragmatic perspective is that the demand for Wi-Fi 7 in humanoid robots is essentially a projection of industrial-grade reliability requirements onto a new terminal product category. Regardless of whether the terminal form is a robot, AGV, or CNC machine tool, the requirements for MLO (Mean Allocation), interference resistance, and multi-wireless coexistence are the same. While the value spillover from Oufexin's accumulation in this area may not be limited to the humanoid robot category alone, it cannot be expected that humanoid robots alone will support significant revenue growth. Key findings: The penetration of Wi-Fi 7 in industrial settings is a gradual process, with humanoid robots representing a new source of demand rather than the sole driving force. For module manufacturers, the strategic significance of humanoid robots lies in providing a real-world scenario to validate industrial-grade reliability design capabilities, rather than a short-term window for mass production. Conclusion For the Wi-Fi industry, this calibration manifests in: a shift from pursuing peak speeds to ensuring link determinism, from a consumer-grade cost-oriented approach to an industrial-grade reliability-oriented approach, and from single-device connectivity to the coexistence of multiple wireless systems. However, its actual position in the humanoid robot market will depend on three verifiable variables: the deployment timeline driven by the Arm platform, the measured stability of 5GHz MLO under domestic 6GHz unavailability conditions, and whether the modules can win customers with the consistency of industrial-grade quality rather than the technical specifications of prototypes when humanoid robots are truly deployed in large quantities . This chain is long, and Wi-Fi 7 is only one link, but it is precisely the engineering maturity of this link that determines whether the robot's "nervous system" can support the industry expectations described by Dai Jianjun.