乐鑫科技(688018):ESPRESSIF SYSTEMS 2026 Q2 & HALF-YEAR REPORT
原标题:乐鑫科技:Espressif Systems 2026 Q2 & Half-Year Report Message from the CEO 03 Management Report （Continued） Developer-Centric Platform Strategy 20 Developer Community Content 21 About Espressif 07 Facts About Espressif 07 Research and Development 22 Key Technologies 08 Ownership of Securities 24 Risks and Uncertainties 25 Core Competitiveness 09 Consolidated Financial Statements 26 (Unaudited) Business Highlights 10 Consolidated Balance Sheet 26 Key Financial Data (Unaudited) 10 Consolidated Income Statement 28 Key Financial Indicators (Unaudited) 10 Consolidated Cash Flow Statement 29 Financial Analysis 11 Consolidated Cash Flow Statement - Indirect 26 Method Consolidated Statement of Changes in Equity 31 Management Report 13 Industry Overview 13 Further Information 32 Product Positioning 15 Terms and Conditions of the Report 32 Key Products - Chip Matrix 16 List of abbreviations 33 Product Market Tenure 17 Key Products - M5Stack Development Kits 18 B2D2B Business Model 19 Message from the CEOEspressif continued to deliver steady growth in the ?rst half of 2026. The Company generated revenue of RMB 1.48 billion, up 18.9% year on year. Gross margin improved by 4.1 percentage points to 49.3%, while net pro?t attributable to shareholders increased 27.9% to RMB 334 million. Growth momentum strengthened during the period, with second-quarter revenue exceeding RMB 800 million for the ?rst time and gross margin reaching 50.0%, re?ecting continued improvement in business quality. We believe sustained and effective R&D investment is the foundation for product differentiation and long-term pro?tability. In the ?rst half of 2026, R&D investment reached RMB 330 million, representing 22.3% of revenue, with a continued focus on core technologies, software platforms and new product development. We allocate resources based on product execution, customer adoption and long-term value creation, converting our technical capabilities into more competitive products and stronger business performance. Our ?rst-half performance was not driven by a single product cycle, but by a broader product portfolio, a growing developer ecosystem and the supply resilience we have built over time. Together, these factors supported revenue growth, pro?tability improvement and the continued expansion of the markets we serve. Broadening our addressable market The ESP32-P4 is an important step beyond our traditional communications portfolio. It is a ? high-performance microcontroller without an integrated radio, designed for applications requiring compute performance, human-machine interfaces and edge processing. In many customer designs, it can operate alongside one of our connectivity products. This allows us to address a larger share of the system while preserving the modularity customers need. We also continued to expand our low-power wireless portfolio. The ESP32-H4 and ESP32-H21 ? support Bluetooth Low Energy and IEEE 802.15.4, providing a foundation for Zigbee, Thread and Matter applications. The H4 is designed for more demanding battery-powered products requiring dual-core processing and advanced Bluetooth capabilities. The H21 extends the portfolio toward cost- and power-sensitive sensing and control nodes. This portfolio matters because connected homes, buildings and industrial systems increasingly use several communications protocols. Low-power mesh devices frequently operate alongside Wi-Fi gateways and bridges. Supporting these applications through a common development framework can reduce integration effort for customers, shorten development cycles and increase the number of applications that Espressif can serve. The ESP32-S31 began sampling with customers in May and of?cially entered mass production ? in July. The S31 combines our self-developed dual-core RISC-V architecture with SIMD acceleration, multi-protocol connectivity and interfaces for audio, vision and human-machine interaction. It is designed to support embedded AI and signal-processing workloads, including intelligent audio and basic computer-vision applications. To address the growing demand for high-performance AI computing at the edge, we are also ? actively advancing the development of our next-generation AI edge chip based on our self-developed RISC-V IP. Expected to be manufactured using an advanced 12nm process technology, this chip is designed to signi?cantly improve energy ef?ciency and local AI inference capabilities, while enabling more complex algorithm deployment and exceptional power ef?ciency. This development will further strengthen Espressif’s architectural independence and competitive position in edge AI. The signi?cance of these additions is not simply that we are releasing more chips. They expand our addressable market from connectivity-focused microcontrollers into general-purpose processing, higher-performance wireless connectivity and edge intelligence. They also give existing customers more reasons to use Espressif products across multiple functions within the The Long-Term Value of Our Ecosystem For more than a decade, we have invested not only in silicon, but also in the broader infrastructure surrounding our chips, including ESP-IDF, open-source tools, technical documentation, reference designs and our developer community. Many of these investments began long before their ?nancial returns became apparent. This ecosystem reduces the practical cost of adopting Espressif products. Engineers can leverage existing software, examples and community knowledge rather than starting each project from scratch. This enables customers to move more quickly from product evaluation to working designs and increases the likelihood that an initial design win evolves into a long-term customer relationship. AI-assisted software development has the potential to further amplify the value of this advantage. Platforms with comprehensive, well-structured documentation and extensive open-source code are generally easier for both engineers and AI-assisted development tools to understand and utilize. As these tools continue to evolve, we expect our long-term investment in open software and documentation assets to create even greater value. We do not view AI as a replacement for engineering expertise or customer support. AI-generated ?rmware still requires validation for security, reliability and performance. Our goal is to make Espressif one of the most accessible and developer-friendly platforms for building reliable embedded products, whether developers use our tools directly or leverage AI assistance. We are also actively exploring new development paradigms in the AI era. Through initiatives such as ESP-Claw and ESP-Vision, as well as our support for emerging AI development technologies such as MCP, we aim to further lower the barriers between AI and physical devices, enabling developers worldwide to build intelligent applications for the real world at lower cost. At the same time, our chips are becoming increasingly capable of running AI workloads locally. Edge processing can reduce latency, dependence on cloud connectivity, ongoing computing costs and the transmission of sensitive data outside the device. We believe that combining more capable edge computing with a high-productivity software ecosystem will further lower adoption barriers and expand the range of commercially viable applications. Extending our engineering capability The ESP32-E22 provides a further demonstration of the progress in our higher-performance 2×2 MU-MIMO and channel bandwidth of up to 160 MHz, and is supported by an open-source Linux driver. Certi?cation is important because it independently validates standards compliance and interoperability. More broadly, developing a Wi-Fi 6E connectivity co-processor with this level of radio, digital and software complexity demonstrates the increasing scope of our internal R&D capability. This capability will support our continued development toward Wi-Fi 7, which remains an important part of our product roadmap. Our current objective is to reach Wi-Fi 7 capability by mid-2027. Supply and working-capital discipline We continue to plan for tight capacity in the parts of the semiconductor supply chain relevant to our products. We have therefore maintained additional inventory and supply coverage to reduce the risk of interruptions for customers. This strategy carries a cost. It uses working capital, increases inventory exposure and can affect cash conversion if demand or product transitions differ from our expectations. We believe the resilience it provides is valuable, particularly because reliable supply can in?uence customer design decisions that last for several years. We will continue to balance that bene?t against demand visibility, inventory ageing and capital ef?ciency. Looking ahead The ?rst half demonstrated progress on the priorities that matter most to Espressif's long-term value: sustainable revenue growth, improving gross margin, continued R&D investment and expansion into adjacent markets where our technology and ecosystem provide a credible advantage. Our portfolio now extends from ultra-low-power Bluetooth LE and IEEE 802.15.4 devices through general-purpose and AI-capable processing to higher-performance Wi-Fi connectivity. The commercial opportunity is larger as a result, but so is the execution challenge. We must convert new products into durable customer designs, manage operating expenses carefully and maintain disciplined use of working capital. Espressif independently designs and develops its own chip products, with core IPs being self-developed. These chips are 75 equipped with rich features, ensuring that Espressif’s products stand out in the market and avoid the pitfalls of homogenized 50 competition. By controlling the entire design process, Espressif can innovate rapidly and introduce unique functionalities that set 25 our products apart from competitors, providing a distinctive edge in the IoT semiconductor industry. 0 2019-07 2026-06 Full-Stack Engineering Capability Espressif possesses comprehensive engineering development capabilities that span from IP development to complete chip design, operating systems, ?rmware, software frameworks, application solutions, hardware design, edge AI, cloud, and apps. This full-stack engineering expertise allows Espressif to offer highly integrated solutions that meet a wide range of customer needs. The ability to control and optimize every layer of the technology stack is unique in the industry and enables Espressif to deliver superior products and services that are seamlessly integrated and highly ef?cient. Superior Cost-Effective and Stable Support Espressif's products are known for their high performance and low cost, making them competitive in the market. We provide the cost ef?ciency necessary to promote large-scale commercialization in downstream industries. Additionally, we ensure the long-term availability of our products and offer stable and enduring software support, ensuring that customers can rely on their investments for extended periods. This commitment to value and reliability strengthens customer trust. Extensive Developer Community Support Espressif has garnered support from a vast community of professional engineers who are familiar with Espressif's development platform and actively promote its value proposition. This large and engaged community not only aids in the rapid adoption and dissemination of Espressif’s technologies but also contributes to continuous improvement through feedback and shared knowledge. The robust community support ampli?es Espressif's market presence and Business Highlights Key Financial Data (Unaudited) Three Months Ended Six Months Ended CNY Jun 30, 2026 Jun 30, 2025 Jun 30, 2026 Jun 30, 2025 Consolidated Statement of Income Data 832,739,862 687,690,927 1,480,711,294 1,245,540,796 Revenue Gross pro?t 415,932,633 321,088,233 730,254,956 563,035,750 24,787,914 20,891,691 47,935,740 37,911,612 Selling expenses General and administrative expenses 25,665,226 20,014,920 47,402,114 37,868,504Research and development expenses 175,180,569 142,052,489 329,666,051 268,244,538Net income 199,233,110 168,598,092 335,278,541 263,039,765 Net income attributable to Espressif 198,583,901 167,557,823 334,226,057 261,262,177Earnings per share: Basic 1.4314 1.2049 Diluted 1.4314 1.2049 Consolidated Cash Flow Data Net cash provided by operating activities 69,029,210 39,450,437 72,446,512 114,422,825Non-GAAP Adjustment Stock-based compensation 21,540,785 13,481,337 28,066,954 18,790,505Non-GAAP net income attributable to Espressif 220,124,685 181,039,160 362,293,011 280,052,682Jun 30, 2026 Dec 31, 2025 Consolidated Balance Sheet Data 3,737,844,947 3,518,352,994 Working capital Total assets 5,280,396,574 5,047,755,518 Long-term obligations 154,215,210 141,345,125 Total shareholders’ equity 4,737,937,020 4,489,877,928 N.B.: Due to the company's implementation of a capital reserve to equity conversion in 2026, the earnings per share for each reporting period have been recalculated based on the adjusted share count. Key Financial Indicators (Unaudited) Three Months Ended Six Months Ended Jun 30, 2026 Jun 30, 2025 Jun 30, 2026 Jun 30, 2025 49.95% 46.69% 49.32% 45.20% Gross margin R&D-to-sales ratio 21.04% 20.66% 22.26% 21.54% 27.88% 26.69% 26.27% 23.24% EBITDA margin Net income margin 23.93% 24.52% 22.64% 21.12% Financial Analysis Revenue & Margin: First-half revenue reached CNY 1.48 billion, representing an 18.9% YoY increase, with second-quarter revenue growth exceeding 20%. Although global smart home demand moderated to single-digit growth amid a challenging macroeconomic environment and ongoing geopolitical uncertainties, Espressif continued to bene?t from the diversi?cation of its application portfolio. Non-smart home segments—including industrial automation, smart agriculture, energy management, and AI-driven developer and innovation activities—continued to deliver solid growth. Supported by the Company's strong developer ecosystem, broad customer base, and recognized brand position, Espressif continued to demonstrate pricing power in the market. Together with ongoing product mix optimization, these factors contributed to an overall gross margin of 49.3%, remaining at a high level. Revenue by Area: China’s Mainland revenue saw a 31.0% QoQ increase and a 15.9% YoY growth for the second quarter, bringing the ? H1 cumulative YoY growth to 12.2%. ? Overseas revenue grew by 23.5% QoQ and 34.1% YoY in Q2, with an H1 cumulative YoY growth rate of 35.8%. Revenue by Product Category: ? Chips: Revenue grew by 6.6% YoY, re?ecting steady demand from the customer base that prefers direct chip-level design and in-house hardware development. Modules & Development Kits: Revenue grew by 26.8% YoY, as the IoT ecosystem continues to mature. An ? increasing number of customers are choosing pre-certi?ed module solutions to simplify hardware development, shorten product development cycles, and concentrate engineering resources on software, AI, and application innovation. Revenue by Area Revenue by Sales Top-5 Customer in H1 2026 Mode in H1 2026 Concentration Rate 2022 2023 2024 2025 H1 2026 Revenue by Product Category Pro?tability & Resilience: Net income attributable to shareholders reached CNY 334.23 million, representing a 27.9% year-on-year increase, outpacing revenue growth. Despite a softer market environment, the Company continued to demonstrate strong operational resilience through sustained margin expansion, diversi?ed end-market exposure, and continued execution of its long-term growth strategy. R&D investment: R&D expenses increased by 22.9% YoY, primarily driven by higher R&D headcount and employee compensation. The increase re?ects the Company's continued investment in expanding its technology capabilities in response to the industry's evolution from basic IoT connectivity toward intelligent edge computing and Physical AI applications. Cash Flow & Inventory: During the reporting period, memory prices increased and procurement lead times extended. To secure supply and ensure stable production and deliveries, the Company strategically increased inventory reserves. Consequently, inventory rose by CNY 214.20 million compared with the end of 2025, which was the primary reason why net cash generated from operating activities stood at CNY 72.45 million, lower than net pro?t. This deliberate working capital investment is expected to normalize as supply chain conditions stabilize. ROE & Capital: The decline in ROE was a result of capital dilution following the new share issuance completed in Q3 2025, rather than a shift in underlying pro?tability. Key Financial Indicators H1 2025 H1 2026
Industry Overview The rapid development of arti?cial intelligence (AI) technology is propelling the Internet of Things (IoT) industry into a new phase of growth. As generative AI, edge computing, wireless communications, and open-source software ecosystems continue to mature, the foundational logic driving smart terminal development is undergoing a profound shift. The focus of industry competition has gradually expanded from traditional device connectivity to computing power, intelligent interaction capabilities, and software ecosystem capabilities. The convergence of AI and IoT (AIoT) is continuously spawning new application scenarios, while driving the smart terminal industry toward a more open and diversi?ed development stage. 1. AI Drives Smart Terminals from "Connectivity" to "Intelligence" Over the past decade, the development of the IoT industry primarily revolved around device networking, data collection, and remote control, where connectivity served as the most critical foundational capability for smart terminals. With the rapid evolution of large language models (LLMs), edge AI, and multimodal technologies, terminal devices are increasingly equipped with visual perception, voice interaction, local inference, and autonomous decision-making capabilities. Consequently, the industry's focus is shifting from "whether a device is connected" to "what tasks it can accomplish." Concurrently, AI computing architectures are undergoing structural shifts. To satisfy requirements for real-time response, privacy protection, network bandwidth, and power consumption control, an increasing volume of AI inference is migrating from the cloud to the edge. According to IDC forecasts, the global edge computing market is expected to maintain double-digit growth over the coming years, with market size exceeding $380 billion by 2028. Gartner also notes that a vast amount of enterprise data will be generated and processed at the edge in the future, making edge computing an essential component of AI deployment. This trend will continuously drive the demand for edge computing platforms, high-performance MCUs, SoCs, and related software platforms. Meanwhile, the global volume of smart devices continues to experience long-term growth. According to forecasts by TSR, annual shipments in the wireless connectivity chip market will exceed 15 billion units by 2032, with connected endpoints continuously expanding into industrial automation, energy management, smart agriculture, commercial equipment, robotics, and various emerging ?elds. The continuous enhancement of AI capabilities is enabling smart terminals to gradually evolve from simple "connected devices" into a new generation of intelligent devices capable of autonomous perception and smart interaction. 2. Shifting Drivers of Innovation: Developers Emerging as a Core Force Beyond the continuous upgrading of terminal form factors, AI is fundamentally changing the innovation model of the smart hardware industry. In the past, new product R&D relied heavily on large terminal manufacturers, characterized by longer R&D cycles and highly concentrated innovation directions. In recent years, however, the rapid proliferation of LLMs, open-source software, open-source hardware, and AI development tools has signi?cantly lowered the barriers to smart hardware development. An increasing number of startups, academic and research institutions, individual developers, and open-source communities are actively participating in smart terminal innovation, resulting in a more decentralized, diversi?ed, and rapidly iterative industry landscape. According to GitHub's Octoverse report, AI has become one of the fastest-growing open-source technology domains globally, with the number of AI-related open-source projects and developers increasing rapidly. Simultaneously, novel products such as robotics, AI toys, intelligent companion devices, edge vision terminals, wearable devices, and smart commercial terminals are continually emerging. A vast majority of these innovative Industry Overview (Continued) Moving forward, industry innovation will increasingly originate from global developers, startup teams, and open communities, rather than relying solely on a small number of enterprise giants. The strategic importance of development ef?ciency, software ecosystems, and innovative toolchains will continue to elevate. 3. Platform-Based Competition Becomes a Key AIoT Trend Alongside the rapid adoption of AI applications, industry competitive dynamics are evolving. The traditional semiconductor industry competed primarily on chip performance, process nodes, cost ef?ciency, and supply chain capabilities. In contrast, the AI era introduces signi?cantly higher software complexity. Developers today focus not only on hardware performance, but even more so on development tools, software frameworks, AI model support, operating systems, open-source ecosystems, and cross-platform development capabilities. Software ecosystems, open-source collaboration, and developer toolchains have become critical factors in?uencing R&D ef?ciency. An increasing number of enterprises are leveraging open software platforms, uni?ed development frameworks, and comprehensive toolsets to help developers shorten product development cycles and accelerate the commercial rollout of innovations. At the same time, the application boundaries of AI terminals continue to expand. Extending beyond smart homes into robotics, industrial automation, smart energy, smart agriculture, healthcare, and commercial terminals, many emerging applications remain in a phase of rapid exploration with ?uid market boundaries. This implies that future industry growth will stem not only from increased penetration in existing markets, but also from the continuous emergence of new applications and market demands. In summary, arti?cial intelligence is pushing the smart terminal industry into a brand-new phase of development. The ongoing advancement of edge intelligence, multimodal interaction, wireless connectivity, open-source software, and developer ecosystems is collectively shifting AIoT competition from standalone hardware specs to platform- and ecosystem-centric strength. As AI continuously lowers the barriers to innovation, future smart terminal innovation will become increasingly vibrant and application scenarios will continue to expand, opening up long-term strategic opportunities for the entire industry.
Interaction Uni?ed Developer Ecosystem across all segments The Company's product roadmap is structured around four application-oriented capability layers: Basic IoT, Systematic Connectivity, Interactive Devices, and Physical AI & High-Interaction. This roadmap re?ects the evolution of connected devices from simple sensing and communication toward intelligent, autonomous, and interactive systems. Basic IoT focuses on delivering reliable, cost-effective connectivity for large-scale embedded devices. Products such as the ESP32-C series provide highly integrated Wi-Fi and Bluetooth connectivity with optimized power consumption and cost ef?ciency, enabling broad adoption across smart home devices, consumer electronics, industrial control, and other mainstream IoT applications. Building upon basic connectivity, Systematic Connectivity addresses increasingly complex networking environments where multiple wireless technologies operate together. The Company's portfolio supports Wi-Fi, Bluetooth LE, Thread, Zigbee, Matter, and Wi-Fi 6/6E, enabling devices to participate in heterogeneous networks with improved bandwidth, lower latency, stronger reliability, and lower power consumption. This capability allows customers to build scalable and interoperable connected systems rather than isolated devices. Interactive Devices combine connectivity with local intelligence to enable richer human-device interaction. Products including the ESP32-S series integrate AI acceleration, vector instructions, graphics capabilities, and multimedia processing to support voice interaction, vision processing, display control, and other real-time edge AI applications, enabling a new generation of intelligent terminals. Looking further ahead, the Company is expanding toward Physical AI & High-Interaction. Products such as the ESP32-P series provide signi?cantly higher computing performance, richer peripheral interfaces, multimedia capabilities, and support for embedded Linux, forming the computing foundation for next-generation intelligent control systems, robotics, and AI Agent applications. At the software layer, the Company is also investing in technologies such as MCP integration, ESP-Claw, and AI-native development frameworks to help developers more Key Products - Chip Matrix
Wi-Fi 4 Bluetooth Wi-Fi 6 Thread Wi-Fi 6E Zigbee Matter Ultra-Low PowerProduct Market Tenure and stable performance across complex application environments—reliability that cannot be replicated by speed or short-term iteration alone. Long-Tail Value Products within the IoT ecosystem are capable of sustaining revenue contributions over extended periods. Unlike consumer mobile devices, IoT chips typically follow a "Slow Burn, High Retention" model where reliability is the primary purchasing driver. Business Model Business-to-Developer-to-Business (B2D2B) We believe developers are the primary drivers of technological innovation. By serving developers ?rst, we enable the creation of new products, new applications, and ultimately new markets. Our growth model is built on the global developer ecosystem through two complementary layers: Business-to- Developer (B2D) and Business-to-Developer-to-Business (B2D2B). B2D (Business-to-Developer) forms the foundation of our ecosystem. We regard developers as our primary direct users and continuously invest in a comprehensive development platform, including processors, software frameworks, operating systems, development tools, development kits, and technical support. Through our open- source ecosystem, technical community, developer events, and continuous software enhancements, we maintain long-term engagement with developers worldwide. In this sense, B2D resembles a direct-to-consumer (DTC) model, where the "consumer" is the global developer community. B2D2B (Business-to-Developer-to-Business) transforms developer adoption of our platform into enterprise adoption of our technologies. As developers build innovative products on our platform, they bring our technologies into their organizations and industries, driving the development and commercialization of intelligent hardware applications across industrial automation, energy management, smart agriculture, consumer electronics, AI-enabled devices, and many other markets. Together, these two layers form Espressif's Developer Flywheel, where developers create new applications that drive enterprise adoption, while growing enterprise adoption attracts more developers, reinforcing a self-sustaining cycle of developer growth, application innovation, enterprise adoption, and ecosystem expansion.Behind Every Smart Devices is a Developer-Centric Platform StrategyEspressif's long-term development strategy is centered on continuously expanding the developer ecosystem served by its technology platform. As intelligent connected devices evolve, the required skill sets and development work?ows continue to diversify. Accordingly, the Company is broadening its technology portfolio to support an increasingly diverse range of developer communities. Today, Espressif's platforms are widely adopted by Connected Device Developers, who focus on low-power wireless IoT applications, and Smart Interaction Developers, who build products featuring displays, voice, cameras, touch interfaces, and multimodal user experiences. The growing adoption of on-device AI has also expanded the Company's engagement with Edge AI Developers, while increasing demand for higher bandwidth and lower latency has created new opportunities among High-Performance Connectivity Developers developing multimedia, industrial networking, and advanced wireless applications. Looking forward, the Company is making strategic investments to support the next generation of embedded software development. This includes ongoing research and development of Linux-capable platforms and software frameworks intended to serve Linux Application Developers, enabling more sophisticated embedded computing applications. In parallel, the Company is exploring AI-native development frameworks and work?ows, including AI- Developer Community Content @20260630 Cumulative Number of GitHub ESP32 Projects 179 K BOOKS Global Code Hosting Platform Social Networking Site Developer-written books 235,800 196,359 > 300 ESP32 and ESP8266 ESP32 Group Members Covering more than 10 Projects languages 2016-12 2026-06 Daily Increase on GitHub ESP32 Projects 319 Chinese Video Sharing Platform Global Video Sharing Platform 5.17 M 3.76 M 903 K Most Viewed ESP32 Most Viewed ESP32 Most Viewed ESP32 Video (2026) Short (2026) Video (2026) 2016Q4 2026Q2 Always a on Social Media Posted on 13 Jul 2026 Posted on 2 Feb 2026 Posted on 9 Jul 2026 Posted on 26 Jul 2026 ????，从零硬啃机械臂！电机?的是 谱めくりペダル（?作）できた… Researchers built a soft ?oating robot I Built a Tiny World Cup Live Score ?体式步进+驱动，? CAN 总线，主控是 for indoor interaction. Gadget ? ESP32-S3 今回は?幅に?を省くためにGeminiに 助けてもらっちゃいました It uses helium and ?apping ?ns instead ? I built a tiny DIY FIFA World Cup Live of propellers. Score Gadget powered by an ESP32! The result is quiet, lightweight, and safe to touch. In this video, I'll show you how I built a It can follow people, give reminders, compact ESP32-based World Cup live and act as a study buddy. score tracker that displays live football scores, goal alerts, match statistics,…… https://www.bilibili.com/video/ https://x.com/tyuuuytrewq/status/ BV1G9Nw6kEZE/ 2081183509989949858Research and Development R&D Investment As the Company's product roadmap advances from Basic IoT to Systematic Connectivity, Interactive Devices, and ultimately Physical AI & High-Interaction, R&D efforts have expanded beyond wireless connectivity to include embedded computing, multimedia processing, embedded Linux, AI software frameworks, developer tools, and edge AI technologies. The growing complexity of both hardware and software platforms requires broader engineering expertise and deeper cross-disciplinary collaboration. Accordingly, the increase in R&D investment re?ects not only the expansion of product development activities, but also the Company's continued transition from a connectivity semiconductor supplier to a provider of integrated computing, connectivity, and software platforms for intelligent edge devices. Fiscal Year CNY June 30, 2026 June 30, 2025 change (%) R&D Expenses 329,666,051 268,244,538 22.90 R&D-to-sales ratio 22.26% 21.54%List of Intellectual Property Rights Applied and Granted H1 2026 Cumulative number Applied for Granted Applied for Granted Invention patents 15 4 196 114 1 1 41 39 Utility model patents Design patents 5 4 20 15 Software copyright 0 0 27 27 Others 9 3 87 44 30 12 371 239 Total N.B.: The number of intellectual property rights that have been “applied for" does not include the rejected applications in China, or the PCT patent applications not R&D Employees We continue to enrich our engineering organization with talent spanning embedded systems, wireless connectivity, AI, operating systems, Linux, software frameworks, graphics, multimedia, cloud services, and developer tools. This multidisciplinary expertise enables us to support a broader developer community—from embedded engineers and product designers to AI application developers and software engineers—while lowering the barriers to building intelligent connected products. We view hardware and software as a uni?ed technology stack rather than separate disciplines. Alongside our semiconductor roadmap, we continue to invest in AI-native software platforms, operating systems, development frameworks, toolchains, middleware, and cloud services. Our objective is to provide developers with an integrated, end-to-end development experience that accelerates innovation from prototype to mass production. At the same time, AI is transforming the way we build products internally. Across software development, hardware design, testing, documentation, and knowledge management, our engineering teams are increasingly adopting AI- assisted tools to automate repetitive work, improve collaboration, and shorten development cycles. Rather than replacing engineers, these tools enable them to dedicate more time to system architecture, product innovation, and solving complex engineering challenges. Looking ahead, we will continue to invest in both engineering talent and AI-enabled productivity. We believe that the combination of a world-class engineering team, an expanding software platform, and continuously improving development ef?ciency will strengthen our ability to serve developers worldwide and reinforce the long-term competitiveness of the Espressif ecosystem. Fiscal Year June 30, 2026 June 30, 2025 R&D employees 653 589 R&D employees as a percentage of total employees 68.45% 70.12% Total compensation of R&D employees (CNY in thousands) 239,268 206,006 373 361 Average compensation of R&D employees (CNY in thousands) R&D Expenses R&D Employees 228.2%8.2% 226.6.5%5% 7777..4%4% 776.6.1%1% 224.4.4%4% 7711.8%.8% 7700..0%0% 2233..5%5% 68.68.5%5% 222.2.3%3% 603 653 629 490 553 404 4842022 2023 2024 2025 H1 2026 2022 2023 2024 2025 H1 2026 Risks and Uncertainties Financial Risk The Company performs periodic credit evaluations of its customers’ ?nancial condition and generally requires of its customers no collateral. The Company provides an allowance for expected credit losses, based on the net amount expected to be collected on such receivables. Losses have not been signi?cant for any of the periods presented in this report. R&D Risk Espressif’s research and development strategy is focused on leveraging new technologies for the creation of innovative AIoT products. Any delays or changes in the development of these technologies by our industry partners, or a failure of our products to achieve market acceptance, could compromise our competitive position. Nordic, Silicon Labs, and others. Intense competition from current players may reduce our product sales and market share. Suppliers A signi?cant portion of the Company’s products is fabricated by the Taiwan Semiconductor Manufacturing Company Limited (TSMC). The inability of TSMC to deliver wafers to the Company in a timely manner could impact the production of the Company’s products for a certain period of time, which could have an adverse effect on the Company’s business, ?nancial condition, results of operations and cash ?ow. Customers The Company sells directly to end customers, distributors, solution providers and contract manufacturers. Our customers are a mix of several big customers and numerous small customers. The concentration ratio for our top-?ve customers was 20.32% in H1 2026. Consolidated Financial Statements Consolidated Balance Sheet (unaudited)Consolidated Balance Sheet (unaudited)Consolidated Income Statement (unaudited)Consolidated Cash Flow Statement (unaudited)Consolidated Cash Flow Statement - Indirect Method (unaudited)Consolidated Statement of Changes in Equity (unaudited) F I S S Y S S E T Declaration Further Information Terms and Conditions of the Report Responsibility Statement The Board of Directors, and the executive management of the Company warrant that the contents of this report are true, accurate and complete, and do not contain any false information, misleading statements or material omissions, severally and jointly accepting any legal responsibility thereof. Shanghai, 30 July 2026 Espressif Systems Executive Management Board of Directors Teo Swee Ann Teo Swee Ann Founder and Chairman Founder and CEO Ng Pei Chi Wang Jue Information Technology Director Board Secretary Wang Jue Shao Jingbo Board Secretary Financial Director Teo Teck Leong Shareholder-elected Chen Myn Independent Director Lee Kian Soon Independent Director Leong Foo Leng Independent Director Forward-looking Statements This report contains forward-looking statements and/or assessments about the business, ?nancial condition, performance and strategy of the Espressif Group. These statements and/or assessments are based on assumptions and management expectations resting upon currently available information and current estimates. These are subject to a multitude of uncertainties and risks, many of which are partially or entirely beyond Espressif’s control. Espressif’s actual （未完）  |