Core Insights - Huawei has launched the world's strongest computing supernodes and clusters at the Huawei Connect 2025 event, emphasizing the critical role of computing power in artificial intelligence and its importance to China's AI development [1][2] - The newly released supernodes, Atlas 950 SuperPoD and Atlas 960 SuperPoD, support 8192 and 15488 Ascend cards respectively, leading in key metrics such as card scale, total computing power, memory capacity, and interconnect bandwidth [1] - Huawei has introduced the world's first general-purpose computing supernode, TaiShan 950 SuperPoD, which aims to replace various large and small machines, including Exadata database machines, in multiple application scenarios [2] Computing Power and AI Development - The company is confident in providing sustainable and ample computing power for the long-term rapid development of artificial intelligence, leveraging its supernode and cluster technologies [2] - Huawei has developed a unified interconnect protocol, Lingqu (UnifiedBus), to address the challenges of large-scale supernode interconnect technology, and plans to open the Lingqu 2.0 technical specifications for industry partners to develop related products [2] Future Outlook - The company aims to continuously meet the growing demand for computing power based on the Lingqu-based supernodes and clusters, promoting the ongoing development of artificial intelligence and creating greater value [2]

Core Insights - The recent WAIC highlighted the growing interest in supernodes, with companies like Huawei, ZTE, and H3C showcasing their advancements in this technology [3][4][5] - Supernodes are essential for managing large-scale AI models, enabling efficient resource utilization and high-performance computing [3][4][5] - The shift from traditional AI servers to supernode architectures is driven by the increasing complexity and size of AI models, which now reach trillions of parameters [4][5][9] Group 1: Supernode Technology - Supernodes integrate computing resources to create low-latency, high-bandwidth computing entities, enhancing the efficiency of AI model training and inference [3][4] - The technology allows for performance improvements even when individual chip manufacturing processes are limited, making it a crucial development in the industry [4][9] - Companies are exploring both horizontal (scale out) and vertical (scale up) expansion strategies to optimize supernode performance [5][9] Group 2: Market Dynamics - Domestic AI chip manufacturers are increasing their market share in AI servers, with the proportion of externally sourced chips expected to drop from 63% to 49% this year [10] - Companies like墨芯人工智能 are adopting strategies that focus on specific AI applications, such as inference optimization, to compete with established players like NVIDIA [10][11] - The competitive landscape is shifting, with firms like云天励飞 and后摩智能 targeting niche markets in edge computing and AI inference, avoiding direct competition with larger chip manufacturers [11][12][13] Group 3: Technological Innovations - The introduction of optical interconnects in supernode technology is a significant advancement, providing high bandwidth and low latency for AI workloads [6][9] - Companies are developing solutions that leverage optical communication to enhance the performance of AI chip clusters, addressing the limitations of traditional electrical interconnects [6][9] - The focus on sparse computing techniques allows for lower manufacturing process requirements, enabling more efficient AI model computations [11][12]

Core Insights - The World Artificial Intelligence Conference should not be viewed merely as an exhibition, with Huawei's 384 super nodes being the most eye-catching highlight [1] - In the context of global computing power chip limitations and an accelerating arms race, super node technology has become a key to breaking through current challenges [1] Company Highlights - Huawei showcased its 384 super nodes at the conference, emphasizing its technological prowess and innovation in the AI sector [1] Industry Trends - The event reflects the growing importance of super node technology in addressing the limitations of computing power chips globally [1] - The acceleration of the arms race in technology indicates a competitive landscape where advancements in AI and computing power are critical [1]

Core Viewpoint - The emergence of Huawei's Ascend CLoudMatrix 384 supernode, which surpasses NVIDIA's flagship NVL72 system by 70% in computing power, signifies a shift in the AI computing landscape from single-point breakthroughs to system-level innovations, driven by the need to overcome traditional computing limitations under U.S. sanctions [1][6][29]. Group 1: AI Computing Landscape - The AI computing race is transitioning from hardware-centric approaches to architecture redefinition, with Huawei's innovations highlighting a unique path for China's system-level advancements [1][6]. - Huang Renxun, CEO of NVIDIA, has expressed increasing anxiety regarding China's rapid advancements in AI technology, emphasizing the impossibility of halting China's progress in this field [2][5][9]. Group 2: Huawei's Technological Advancements - Huawei's Ascend CLoudMatrix 384 supernode utilizes domestic Ascend chips and achieves a total computing power of 300 PFlops, significantly exceeding NVIDIA's NVL72 system [1][6][14]. - The architecture of the Ascend CLoudMatrix 384 supernode is based on a "fully equal architecture," which enhances communication efficiency and overcomes traditional bottlenecks such as the "memory wall" and "communication wall" [1][18][20]. Group 3: Competitive Dynamics - The U.S. government's sanctions have prompted NVIDIA to incur a $5.5 billion inventory loss, while simultaneously highlighting the importance of the Chinese market for NVIDIA's future [5][6]. - Huang Renxun acknowledges that China's advancements in AI technology could lead to a significant reduction in NVIDIA's market share in China, which has dropped from 95% to 50% in recent years [9][22]. Group 4: System-Level Innovations - The Ascend CLoudMatrix 384 supernode's design allows for the integration of thousands of cards, enabling it to support larger models and enhance training efficiency [1][6][14]. - The use of optical communication technology in the Ascend CLoudMatrix 384 supernode allows for high bandwidth and low latency, which is crucial for large-scale AI model training [20][21]. Group 5: Future Implications - The successful deployment of the Ascend CLoudMatrix 384 supernode and its ability to train large models like the Pangu Ultra MoE model demonstrates the potential for domestic AI infrastructure to achieve self-sufficiency [26][29]. - The emergence of Huawei's technology provides a viable alternative to NVIDIA's offerings, potentially reshaping the competitive landscape in the AI industry [22][29].

Core Viewpoint - The article discusses the competitive landscape of AI computing power, highlighting Huawei's CloudMatrix 384 super node technology as a significant advancement in the face of U.S. export controls on advanced chips, particularly targeting Huawei's Ascend AI chips [2][4][19]. Group 1: U.S. Export Controls and Market Dynamics - On May 13, the U.S. Department of Commerce announced a global ban on Huawei's Ascend AI chips, expanding the ban to all advanced computing ICs from China [2]. - Despite these restrictions, the U.S. tech industry, particularly NVIDIA, is still eager to tap into the Chinese AI market, as evidenced by NVIDIA's announcement of a large order from Saudi Arabia on the same day the ban was issued [2][3]. - The performance degradation of NVIDIA's H20 GPU, which will see a reduction in INT8 precision computing power by over 60%, raises questions about the viability of continued sales to China [3][4]. Group 2: Huawei's Technological Advancements - Huawei's CloudMatrix 384 super node technology can aggregate 384 Ascend computing cards to achieve a computing power of 300 PFlops, rivaling the performance of NVIDIA's H100 GPU [4][13]. - The technology features a new high-speed bus network that enhances inter-card bandwidth by over 10 times, allowing for near-lossless data flow between cards, thus improving training efficiency to nearly 90% of NVIDIA's single-card performance [13][14]. - The CloudMatrix 384 super node is designed to support large-scale expert parallelism, making it compatible with current mainstream models like DeepSeek and GPT [14]. Group 3: Competitive Landscape and Industry Trends - The super node technology represents a critical solution to global AI computing power challenges, with various companies, including NVIDIA and AMD, developing their own versions of super node architectures [15][16]. - Huawei's CloudMatrix 384 is currently the only commercially available large-scale super node cluster globally, having been deployed in Wuhu data center [17]. - The article emphasizes the importance of a comprehensive AI infrastructure that integrates hardware, software, and services, positioning Huawei as a leader in this domain [21][25]. Group 4: Broader Implications and Future Outlook - The ongoing U.S. technology blockade has inadvertently accelerated China's advancements in chip manufacturing and AI technologies, as noted by Bill Gates [19][21]. - The article concludes that modern AI competition is not just about individual chips or models but requires a holistic approach that encompasses a complete ecosystem of hardware and software solutions [21][24].

Group 1: Cloud Computing and AI Infrastructure - The outlook for capital expenditure from cloud giants is positive, with significant growth observed in core companies such as optical modules and copper connections in Q1 [1][2] - Google's Q1 2025 report indicated a 28% year-on-year increase in cloud revenue, with capital expenditure reaching $17.2 billion (up 43% year-on-year) and an expected annual capital expenditure of $75 billion [1] - Amazon clarified that its recent adjustments in data center leasing are part of routine capacity management and do not indicate a reduction in overall data center construction plans [1] Group 2: Domestic Computing Power and AIDC Sector - The AIDC sector is driven by expectations as new capacity investments take time to convert into revenue, with significant growth in fixed assets and construction projects reported by companies like Runze Technology and Aofei Data [3] - Changes in H20 sales regulations are expected to accelerate the adoption of domestic computing power in the second half of the year, with companies like Ascend, Haiguang, and Cambricon being highlighted as key players [3] Group 3: Optical Communication and Copper Connection - Strong capital expenditure is leading to high certainty in performance for sectors such as optical modules, PCB, copper connections, and power supplies, with companies like Zhongji Xuchuang and Xinyi Sheng reporting substantial profit increases [2] - In the copper connection sector, companies like Wolong Nuclear Materials and Dingtong Technology reported year-on-year profit increases of 39% and 212%, respectively [2] Group 4: Military Information Technology - The recovery of orders in military information technology is notable, with satellite internet projects expected to drive demand for new satellite payloads and ground terminal equipment [4] Group 5: Recommended Stocks - Key stocks to watch in the optical communication sector include Zhongji Xuchuang, Xinyi Sheng, and Shijia Photon [5] - In the copper connection sector, recommended stocks include Wolong Nuclear Materials and Dingtong Technology [6] - For domestic computing power, companies like Cambricon and Haiguang are highlighted, while in military information technology, stocks such as Chengchang Technology and Zhenlei Technology are recommended [6]