Starting September 7, SDMC (00901) has been officially included in the Shanghai-Hong Kong Stock Connect scheme, broadening access for eligible mainland investors to participate in the company. Concurrently, the company has been selected for Google Cloud's proactive Physical AI acceleration program, securing new support for the technical integration, scenario exploration, and product commercialization of its AI Home family agent, Cedar. Understanding these two developments merely as a capital markets event and a routine technology partnership would underestimate their significance. The former expands the company's connection with mainland investors, while the latter places SDMC's long-accumulated capabilities in operator relationships, terminal devices, platform architecture, and global delivery into the real-world testing ground of the next phase of AI industry competition. What warrants closer observation is not simply which models have been integrated, but whether AI can successfully enter the home, a complex environment with multiple devices, multiple family members, and diverse services, and evolve into a service ecosystem that operators can deploy, users can utilize, and that can be continuously iterated upon.
From Android TV/Google TV to Physical AI: A Continuously Expanding Google Ecosystem Partnership with a Replicable Foundation
SDMC's collaborative foundation with the Google ecosystem extends far beyond any single terminal device or cloud service. Founded in 2003, SDMC is the world's largest supplier of Android TV intelligent terminal devices. As a key strategic partner within the Google ecosystem, the company has deepened its collaboration with Google since first obtaining Google Android TV certification in 2017. In 2023, SDMC achieved another milestone by becoming the world's first ODM enterprise to receive Google TV certification for projection products, solidifying its position as an integral part of the Google ecosystem. In the cloud phase, according to publicly available Google Cloud customer case studies, SDMC chose Google Cloud in early 2022 to support its XMediaTV cloud service built on the Google ecosystem, resolving issues related to lengthy traditional local deployments, limited scalability, and high operational costs, while helping operators explore personalized advertising and revenue enhancement opportunities. This trajectory from Android TV/Google TV establishing the home entry point, to cloud services connecting content and operations, and now extending towards AI Home and Physical AI, constitutes a partnership path with a replicable foundation. Its value lies not only in being integrated into the Google technology ecosystem, but also in the company's demonstrated experience in synergistically integrating terminals, cloud platforms, data analytics, content services, and operator delivery systems. Consequently, AI Home is not a lab experiment starting from scratch, but rather an extension of existing digital home capabilities into intelligent interaction, scene understanding, task decision-making, and coordination of physical devices within the home environment, laying the groundwork for replicating Physical AI solutions for global operators in the future.
From Connecting Home Devices to Understanding Home Tasks: AI Home Tackles a More Complex Product Challenge
Physical AI should not be equated with simply adding a chat interface to a device. It emphasizes AI's ability to understand context in real-world environments, utilize tools, execute tasks, and continuously adjust based on feedback. The home scenario is particularly complex: user instructions may be incomplete, device capabilities are distributed across televisions, cameras, gateways, and other smart terminals, and services involve content, security, energy conservation, eldercare, and individual family member preferences. Therefore, the core metric for a home agent should not solely focus on conversational fluency, but also on its ability to accurately identify intent, understand household status, select appropriate devices and services, and complete tasks within the boundaries of safety, privacy, and permissions. SDMC officially disclosed that since launching AI Home in 2024, Cedar's applications have expanded from entertainment and home connectivity to areas like security, energy saving, and eldercare, and have integrated the Google Gemini model to enhance reasoning and natural conversation capabilities. From a third-party technical perspective, Google Cloud's public definition of its AI Agent ecosystem program validates this assessment: AI Agents need to be goal-oriented, possess tool use, reasoning, or planning capabilities, and be deployed on Google Cloud; the program supports partners with guidance from product and engineering teams, access to cutting-edge technology and enablement, and provides go-to-market, co-selling, and exposure resources. This indicates that the competition for intelligent agents is shifting from comparing model capabilities to comparing comprehensive capabilities encompassing models, tools, deployment, industry scenarios, and commercialization. For SDMC, Cedar's value lies not in becoming another general-purpose Q&A bot, but in serving as the synergistic entry point for operator home services: orchestrating family member needs, terminal devices, content services, and operator capabilities within a single manageable system.
Significance of Joining the Acceleration Program: Advancing from Technical Validation to Scenario and Product Validation
Based on available collaboration information, the synergy between SDMC and Google Cloud can be understood on three levels. First is the platform and agent layer. Both parties will focus on leveraging the Google Cloud platform and its AI services to enhance Cedar's understanding of home context, handling of complex tasks, and cross-terminal coordination. For home scenarios, this is closer to a real product than single-point model invocation: after a user presents a request, the system needs to identify the task, call upon content or device capabilities, and manage multi-turn feedback. Second is the application scenario layer. Directions like home security, home entertainment, energy conservation, and eldercare each have different requirements for real-time response, reliability, privacy, and interaction. The value of co-creating scenarios lies in technology being tested within real home environments and operator service processes, ultimately solidifying into deployable, operable, and reusable solutions rather than remaining at the demonstration level. Third is the frontier technology co-development layer. Google Cloud's technical and product teams, cloud resources, and ecosystem resources can help shorten the cycle from technical validation to pilot deployment; SDMC's terminal base, operator customer relationships, and global delivery network provide a real productization entry point for new capabilities. The synergy formed resembles a closed loop of "technology supply, scenario validation, customer piloting, and product iteration".
From a Third-Party Perspective, SDMC's Distinctiveness Lies in Bringing AI into the Home, Not Just Showcasing AI
The first observation comes from Google Cloud customer case studies. These cases reveal not conceptual partnerships, but tangible metrics like deployment timelines, video latency, operational scale, and data analysis. This demonstrates that SDMC's collaborative capabilities have already been validated through cloud-based business operations. This engineering experience is equally critical for AI Home, as home intelligent services require stable long-term operation and cannot be effective only during a product launch event. The second observation stems from Google Cloud's own ecosystem logic. The official AI Agent ecosystem program places development, deployment, adoption, and market entry within a single framework, explicitly emphasizing partner support including engineering assistance, Marketplace presence, co-selling, and exposure resources. Therefore, the value of joining the acceleration program should not be interpreted solely as gaining cloud resources, but also as entering an ecosystem with the potential to drive synergy across technology, products, and commerce. The third observation comes from SDMC's publicly displayed ecosystem layout. The company's officially disclosed next steps include leveraging Android TV/Google TV to upgrade large-screen interaction, combining AI TV boxes and edge intelligence to strengthen Cedar's training, inference, and natural conversation capabilities, developing an AI Camera, and embracing the Google Home ecosystem while supporting the Matter standard. This layout indicates that the product boundaries of AI Home are expanding: from being an "assistant on the TV" to becoming a system covering home entry points, visual perception, device interconnectivity, and home services. Of course, a third-party perspective also necessitates caution. Technology partnerships and ecosystem programs themselves do not automatically translate into revenue, user scale, or profit growth. The market will ultimately focus on three verifiable questions: whether pilots can convert into customer orders, whether Cedar can maintain stability and security in real homes, and whether AI capabilities can generate sustainable service revenue or added value for operators.
The Global Operator Network is Key Support for AI Home's Transition from Showcase to Scale
The difficulty in deploying AI Home lies not only in models, but also in delivery. Network conditions, content copyrights, device standards, privacy requirements, and operator business processes vary significantly across different countries. SDMC has long provided integrated software and hardware solutions to global operators. Public reports indicate that as of the end of June 2026, the company has served over 300 operators and retail brands across more than 80 countries and regions. This foundation positions the company to adapt unified AI capabilities to different terminals, networks, and operator businesses. From a productization pathway perspective, SDMC can leverage its existing terminal devices, the XHome device management platform, the XMediaTV video streaming platform, and the Cedar home AI agent to form a cloud-edge-terminal synergistic combination. The cloud handles model, data, and service orchestration; edge devices undertake low-latency interaction and some local processing; terminals bring AI capabilities to televisions, cameras, and home devices; and Google's ecosystem support helps enhance the efficiency and stability of AI capability integration and service deployment. This architecture is closer to a service form that operators can manage and families can actually use. Therefore, objectively speaking, while it will still take time for AI Home to move from technical solution to scaled application, the existing collaborative foundation with Google is expected to reduce trial-and-error costs, facilitating the faster and more stable entry of relevant capabilities into real home environments.
A New Capital Markets Window and a New Cycle of Industrial Implementation
The Stock Connect inclusion provides SDMC with a new channel to reach investors and increases market attention on the company's business progress and information disclosure quality. Public reports citing the company's 2026 interim results indicate that first-half revenue was approximately RMB 2.095 billion, a year-on-year increase of approximately 36.3%; gross profit was approximately RMB 443 million, up approximately 63.4%; and profit for the period was approximately RMB 181 million, up approximately 75.9%. These operational figures provide a reference for the market to understand the company's fundamentals, but should not be directly equated with the future commercialization outcomes of AI Home. A more accurate framework for observation is to view SDMC along two parallel tracks: one is the expansion of capital market connectivity, which requires the company to continuously enhance transparency and verifiability; the other is the deepening synergy within the Google ecosystem, which requires AI Home to achieve phased results in technology, scenarios, customers, and revenue. Joining the Physical AI acceleration program is an important milestone on the latter track, but not the finish line. As relevant capabilities gradually materialize, the market will further observe whether sustainable business value can be generated.
Conclusion: From Ecosystem Collaboration to Real Homes, AI Home Enters Deeper Validation Waters
From Stock Connect inclusion to Physical AI, SDMC is simultaneously facing dual scrutiny from both capital markets and the industrial market. Stock Connect has widened the company's window of market attention, while the Google ecosystem collaboration provides AI Home with synergistic support spanning cloud infrastructure, AI models, terminals, and home connectivity. The more profound shift lies in the fact that competition in home AI is moving from "who can converse" to "who can understand the home, coordinate devices, deliver stable services, and sustain operations". The next phase of collaboration between SDMC and the Google ecosystem will drive AI Home's evolution from on-device intelligence towards more proactive, operable, and scalable cloud-edge collaborative services, which is also the key to the market's understanding of SDMC's long-term value for AI Home.