Core Viewpoint - Ruijie Networks (301165) announced on October 23 that its supernode technology is designed to meet high bandwidth interconnection needs within cabinets, with mature solutions based on 400G and 800G switching chips, supporting interconnection and large-scale deployment of dozens to hundreds of GPU cards [1] Group 1: Technology and Market Opportunities - The construction of supernodes presents new market opportunities for switch manufacturers, highlighting the increasing importance of network nodes and their closer integration with computing nodes to enhance overall efficiency [1] - The company leverages its expertise in Ethernet interconnection and optical communication (LPO, NPO, CPO) to co-create solutions with clients, indicating a collaborative approach to innovation [1] - There are already practical cases demonstrating the market potential of the supernode technology, suggesting a positive outlook for future developments [1]

Summary of Ruijie Networks Conference Call Company Overview - **Company**: Ruijie Networks - **Date**: October 21, 2025 Key Financial Metrics - **Revenue**: 106.1 billion CNY for the first three quarters, up 27.5% year-on-year [2][3] - **Net Profit**: 6.8 billion CNY, up 65% year-on-year [2][3] - **Operating Cash Flow**: Approximately 2.8 billion CNY, nearly a 200% increase [2][3] - **Q3 Revenue**: 40.3 billion CNY, up 6.8% year-on-year [4] - **Q3 Net Profit**: 2.3 billion CNY, roughly flat compared to the previous year [4] Revenue Breakdown - **Domestic Revenue**: 88.4 billion CNY, up 23.9% [7] - **Overseas Revenue**: 18.36 billion CNY, up 48% [7] - **Direct Sales Revenue**: 61.2 billion CNY, up 52%, accounting for 57% of total revenue [8] - **Distributor Sales Revenue**: 45.5 billion CNY, up only 5%, accounting for 43% of total revenue [8] Product Performance - **Network Equipment Revenue**: 92.68 billion CNY, up 37.9% [6] - **Network Security Products Revenue**: 3.4 billion CNY, up 11.7% [6] - **Cloud Desktop Revenue**: 3.7 billion CNY, down 3.8% [6] Market Trends and Insights - **Data Center Market**: Identified as the main battleground for the year, with significant projects won from leading internet companies [9][10] - **Internet Industry Data Center Revenue**: Approximately 52 billion CNY, with 400G products accounting for 34 billion CNY and 800G products exceeding 7 billion CNY [9][22] - **800G Growth Expectation**: Anticipated to double by 2026, although it will not become the mainstream product alone [16][22] Strategic Initiatives - **Focus on Product Innovation**: The company has enhanced market competitiveness through product innovation, particularly in response to rapid AI development [3] - **Overseas Market Strategy**: Continued growth in overseas markets, particularly in Southeast Asia, Middle East, and parts of Europe [17][18] Industry Positioning - **Market Share**: Ranked first in Ethernet switch market share and third in data center switch market share [14] - **Competitive Landscape**: The company is leveraging its strengths in Ethernet interconnection and optical communication to create solutions in the emerging supernode technology space [20][21] Future Outlook - **Q4 and 2026 Expectations**: Optimistic about Q4 performance and 2026 market demand, driven by major players like Alibaba and ByteDance planning significant capital expenditures [15][33] - **Supernode Technology**: Expected to see larger scale applications in 2026, with ongoing developments in GPU interconnection solutions [26][30] Additional Insights - **Cost Management**: The increase in expense ratio is attributed to early business progress, but overall financial data remains aligned with business performance [23][28] - **Domestic Chip Localization**: Progress in domestic chip production is noted, particularly in high-speed products, although challenges remain in the internet and operator application scenarios [24][25] This summary encapsulates the key points from the conference call, highlighting Ruijie Networks' financial performance, market strategies, product developments, and future outlook.

Key Market News - The Ministry of Industry and Information Technology and eight other departments issued the "Work Plan for Stable Growth in the Nonferrous Metal Industry (2025-2026)", projecting an average annual growth of about 5% in the added value of the nonferrous metal industry from 2025 to 2026, with a 1.5% average annual growth in the production of ten nonferrous metals [1] Industry Insights - The DRAM and NAND flash memory markets are facing more severe shortages than previously predicted, with prices expected to rise further in Q4 2025 and 2026. Major companies like SanDisk, Micron, Samsung, and Western Digital have announced price increases, and ADATA has stopped quoting DDR4 prices while prioritizing DDR5 and NAND flash for key customers. NAND flash prices have seen a rise of about 10% recently, indicating a price increase trend in the market [2][3] - The demand for enterprise-level storage products remains high, with significant orders from large cloud service providers shifting NAND supply from consumer to enterprise markets. The DRAM price index has increased by approximately 72% over the past six months, benefiting companies in the storage industry [3] - The AICC 2025 AI Computing Conference in Beijing saw the launch of the "Beijing Plan" aimed at developing industry-specific intelligent systems through collaboration among system manufacturers, large model developers, and application innovators. The SuperPod technology is highlighted as a core innovation in AI computing infrastructure, enhancing performance and efficiency [4] - A research team led by Tsinghua University has made significant advancements in polymer electrolytes for lithium batteries, proposing a new strategy that enhances the physical contact and ionic conductivity of solid-state interfaces. This development is expected to support the commercialization of high-safety, high-energy-density solid-state lithium batteries [5] Stock Movements - Several companies have announced plans for share reductions by major shareholders, including Jingong Intelligent, Weiteng Electric, and Jinhaitong, with reductions ranging from 1% to 3% of total share capital. These reductions are primarily due to personal funding needs [6][7][8]

Group 1 - The AICC2025 Artificial Intelligence Computing Conference was held in Haidian District, Beijing, from September 26 to 27, where over 30 companies, including China Mobile and Inspur Information, jointly released the "Beijing Plan" for intelligent computing applications [1] - The "Beijing Plan" aims to develop a supernode innovation consortium by integrating system vendors, large model developers, and application innovation companies to create industry-specific intelligent entities, promoting the concept of "national core, national connection, national use" [1] - Supernode technology is identified as a core innovation direction for AI computing infrastructure, designed to overcome performance and efficiency bottlenecks of traditional clusters through system-level architecture reconstruction [1] Group 2 - Supernode technology supports trillion-parameter model training and inference with higher energy efficiency and lower latency, facilitating the transition of AI from kilocalories to megacalories [1] - Future advancements in technologies such as optical interconnects and liquid cooling are expected to further reduce computing costs and accelerate the implementation of industry intelligence [1] - Companies mentioned in relation to this development include Feiling Kesi and SMIC [1]

Core Viewpoint - The article provides a detailed comparison of Alibaba's super node with NVIDIA's NVL72 and Huawei's CM384, focusing on GPU count, interconnect technology, power consumption, and ecosystem compatibility. Group 1: GPU Count - Alibaba's super node, known as "Panjun," utilizes a configuration of 128 GPUs, with each of the 16 computing nodes containing 4 self-developed GPUs, totaling 16 x 4 x 2 = 128 GPUs [4] - In contrast, Huawei's CM384 includes 384 Ascend 910C chips, while NVIDIA's NVL72 consists of 72 GPUs [7] Group 2: Interconnect Technology - NVIDIA's NVL72 employs a cable tray interconnect method using NVLink proprietary protocol [8] - Huawei's CM384 also uses cable connections between multiple racks [10] - Alibaba's super node features an orthogonal interconnect without a backplane, allowing for direct connections between computing and switch nodes, reducing signal transmission loss [12][14] Group 3: Power and Optical Connections - NVIDIA's NVL72 uses copper for scale-up connections, while Huawei's CM384 employs optical interconnects, leading to higher costs and power consumption [15] - Alibaba's super node uses electrical interconnects for internal scale-up, with some connections made via PCB and copper cables, while optical interconnects are used between two ALink switches [18][19] Group 4: Parameter Comparison - Key performance metrics show that NVIDIA's GB200 NVL72 has a BF16 dense TFLOPS of 2,500, while Huawei's CM384 has 780, indicating a significant performance gap [21] - The HBM capacity for NVIDIA's GB200 is 192 GB compared to Huawei's 128 GB, and the scale-up bandwidth for NVIDIA is 7,200 Gb/s while Huawei's is 2,800 Gb/s [21] Group 5: Ecosystem Compatibility - Alibaba claims compatibility with multiple GPU/ASICs, provided they support the ALink protocol, which may pose challenges as major manufacturers are reluctant to adopt proprietary protocols [23] - Alibaba's GPUs are compatible with CUDA, providing a competitive advantage in the current market [24] Group 6: Supply Chain Insights - In the AI and general server integration market, Inspur holds a 33%-35% market share, while Huawei's share is 23% [33] - For liquid cooling, Haikang and Invec are key players, each holding 30%-40% of the market [35] - In the PCB sector, the number of layers has increased to 24-30, with low-loss materials making up over 60% of the composition, significantly increasing the value of single-card PCBs [36]

Market News - US stock indices closed higher, with the Dow Jones up 0.27%, Nasdaq up 0.94%, and S&P 500 up 0.48% [1] - International precious metal futures generally declined, with COMEX gold futures down 1.07% at $3678.2 per ounce and COMEX silver futures down 0.12% at $42.1 per ounce [1] - International oil prices slightly decreased, with WTI crude oil down 0.61% at $63.31 per barrel and Brent crude oil down 0.71% at $66.98 per barrel [1] - European stock indices all rose, with Germany's DAX up 1.35%, France's CAC40 up 0.87%, and the UK's FTSE 100 up 0.21% [1] Industry Insights - Huawei's Connect 2025 conference highlighted the importance of supernodes in AI infrastructure, with over 300 CloudMatrix 384 supernodes deployed [2] - Huawei plans to launch the Atlas 950 SuperPoD with a computing power of 8192 cards by Q4 2026 and the Atlas 960 SuperPoD with 15488 cards by Q4 2027 [2] - The supernode technology aims to overcome traditional cluster performance bottlenecks, supporting trillion-parameter model training and inference [3] - Huawei's self-developed low-cost HBM (High Bandwidth Memory) is expected to double computing power annually, enhancing AI server capabilities [4] - A new hydrogen negative ion prototype battery developed by the Dalian Institute of Chemical Physics could significantly impact large-scale energy storage and hydrogen applications [5] - The hydrogen energy market is projected to reach 4 trillion yuan by 2030, driven by technological breakthroughs and policy support [6]

Group 1: Computing Power - Huawei has launched the world's strongest computing power supernode and cluster, introducing the TaiShan 950 SuperPoD, which can replace various large and small machines as well as Exadata database integrated machines [1] - The supernode technology is seen as a core innovation direction for AI computing infrastructure, breaking through traditional cluster performance and efficiency bottlenecks [1] - The technology supports trillion-parameter model training and inference with higher energy efficiency and lower latency, pushing AI from the kilowatt to the megawatt era [1] Group 2: Hydrogen Energy - The Chinese Academy of Sciences has developed the world's first hydrogen negative ion prototype battery, marking a significant breakthrough in solid-state ion battery materials [2] - This new battery technology demonstrates the feasibility of hydrogen negative ions as energy carriers and opens new possibilities for hydrogen energy applications in energy storage [2] - The hydrogen energy market is expected to reach a scale of 4 trillion yuan by 2030, driven by technological breakthroughs and policy incentives [2] Group 3: Short Video Market - The overseas short video app market saw approximately 120.1 million downloads in August, with estimated in-app revenue reaching $196.4 million, marking a significant increase from July [3] - The total revenue for the overseas short video market has surpassed $1.088 billion in the first half of the year, with projections for annual revenue to reach $2.473 billion [3] - By 2027, the overseas short video market is expected to grow to $18.7 billion, potentially surpassing the domestic market [3] Group 4: Macro and Industry News - The China Charging Alliance reported an 88.5% year-on-year increase in electric vehicle charging infrastructure, with 4.53 million new installations from January to August 2025 [4] - China has built approximately 4.6 million 5G base stations, maintaining its position as the global leader in new energy vehicle production and sales for ten consecutive years [5] - The Ministry of Culture and Tourism has launched a three-year action plan to boost cultural and tourism consumption, with over 330 million yuan in subsidies planned [5]

Core Viewpoint - Huawei has introduced an innovative super node architecture aimed at building a robust all-scenario computing foundation, emphasizing open-source and hardware openness to foster industry collaboration and innovation [1][10]. Group 1: Super Node Architecture - The super node architecture, based on the Lingqu interconnection protocol, allows multiple physical machines to be deeply interconnected, enabling them to function as a single logical unit for learning, reasoning, and efficient computing [1][2]. - This architecture addresses the challenges of traditional server stacking, which can lead to lower computing efficiency and frequent interruptions during training as cluster sizes increase [1][2]. Group 2: Product Launches - Huawei has launched several new products based on the super node architecture, including the Atlas 950 SuperPoD, Atlas 850 and Atlas 860 AI servers, Atlas 350 AI accelerator cards, and the TaiShan 950 SuperPoD [4][5]. - The Atlas 950 SuperPoD is designed for large-scale AI computing tasks, featuring innovations such as zero-cable electrical interconnection and enhanced liquid cooling reliability [4]. - The Atlas 850 is the first enterprise-grade air-cooled AI super node server, capable of forming a super node cluster with up to 128 units and 1024 cards [4]. Group 3: Performance Enhancements - The Atlas 350 accelerator card, utilizing the Ascend 950PR chip, offers a 2x increase in vector computing power and a 2.5x performance boost in recommendation inference scenarios [5]. - The TaiShan 950 SuperPoD features ultra-low latency of 370 nanoseconds and a bandwidth of 2.8T, significantly enhancing performance in database and virtual machine migration scenarios [5]. Group 4: Open Collaboration - Huawei is committed to open collaboration by sharing super node technology with the industry, allowing partners to develop products based on the Lingqu protocol and super node reference architecture [6]. - The company has opened its super node hardware, including NPU modules and AI accelerator cards, to facilitate incremental development by customers and partners [6]. Group 5: Software Innovation - Huawei is also focusing on software openness, with plans to open-source the Lingqu operating system components and support various open-source communities, accelerating developer innovation [9]. - The Ascend CANN and Mind series components will be open-sourced, prioritizing support for popular frameworks like PyTorch, enhancing flexibility for developers [9].

Core Viewpoint - Huawei has unveiled its latest advancements in computing power and chip technology, emphasizing the importance of computing power for artificial intelligence (AI) and its critical role in China's AI development [1][4]. Group 1: Super Nodes and Clusters - Huawei introduced the Atlas 950 SuperPoD and Atlas 960 SuperPoD, which support 8192 and 15488 Ascend cards respectively, leading in key metrics such as card scale, total computing power, memory capacity, and interconnect bandwidth [3]. - The Atlas 950 SuperCluster and Atlas 960 SuperCluster were also launched, boasting computing power scales exceeding 500,000 cards and reaching one million cards, making them the strongest computing clusters globally [4]. - Huawei has developed a unified interconnect protocol called Lingqu, which will be open-sourced to encourage industry partners to create related products and components, fostering an open ecosystem [4]. Group 2: Ascend Chip Evolution - Huawei's Ascend chip roadmap includes the 950 series, 960 series, and 970 series, with the 950PR set to launch in Q1 2026 and the 970 series expected in Q4 2028 [6]. - The chips will cover various fields including smartphones, communications, servers, and AI, with a focus on full-stack self-research [6]. Group 3: AI Solutions for SMEs - Huawei has launched the "4+10+N" intelligent solution for small and medium-sized enterprises (SMEs) to facilitate their transition to intelligent operations [8][9]. - The "4" represents four core scenarios: intelligent office, intelligent commerce, intelligent education, and intelligent healthcare, while "10" refers to ten one-stop scenario solutions [10]. - The initiative aims to provide simple and effective intelligent solutions to SMEs, addressing their challenges in technology and talent [9][10].

Core Viewpoint - Huawei has introduced an innovative super node architecture aimed at redefining large-scale effective computing power, emphasizing open-source and hardware openness to foster industry collaboration and innovation [2][3]. Group 1: Super Node Architecture - The super node architecture allows multiple physical machines to be deeply interconnected, enabling them to function as a single logical unit for learning, reasoning, and thinking [2]. - This architecture is designed to meet the computing needs of large data centers, enterprise-level data centers, and small workstations across various industries [2]. - Key features of the super node architecture include resource pooling, scalable expansion, and reliable performance, facilitating high bandwidth and low latency interconnections for computing and storage units [2]. Group 2: New Product Launch - Huawei has launched several new products based on the super node architecture, including the AI super node Atlas 950 SuperPoD, enterprise-level AI super node servers Atlas 850 and Atlas 860, AI next-generation cards Atlas 350, and the first universal super node Taishan 950 SuperPoD [2]. - These products are designed to enhance the capabilities of data centers and support a wide range of computing scenarios [2]. Group 3: Open Source Commitment - Huawei is fully opening its super node technology to share the technological benefits with the industry, promoting inclusive and collaborative innovation [3]. - The operating system components of the Lingqu protocol will be open-sourced, with code being integrated into various upstream open-source communities such as openEuler [3].