Investment Rating - The report does not explicitly state an investment rating for the industry or specific companies involved in the AI sector Core Insights - The WAIC 2025 showcased unprecedented public interest in AI, with significant participation from over 800 companies and more than 3000 cutting-edge exhibits, marking the largest scale in its history [1][7] - The focus of the conference included key themes such as large model applications, computing power infrastructure, AI for science, and the integration of AI in various sectors including finance and healthcare [2][7] - Investment opportunities are shifting from individual GPU cards to full server racks, emphasizing the importance of computing power, model optimization, and AI infrastructure [8][15] Summary by Sections Core Technologies - The emphasis has shifted from individual chip performance to comprehensive solutions, with discussions focusing on efficient utilization of fragmented computing resources and low-power designs [2][12] - Major companies like Huawei and ZTE showcased advanced computing solutions, highlighting the integration of AI chips and servers into practical applications [3][4][5] Industry Applications - WAIC 2025 demonstrated a deeper commercialization of AI, with applications spanning robotics, smart glasses, and healthcare, reflecting a significant increase in participation from state-owned enterprises and international companies [7][13] - The event achieved record-breaking ticket sales and attendance, indicating a growing public enthusiasm for AI technologies [7][13] Smart Devices - The focus on embodied intelligence was evident, with dynamic usage scenarios presented through interactive exhibits that showcased advancements in humanoid robotics and smart glasses [8][14] - Notable product launches included Alibaba's Quark AI Glasses and China Telecom's Tianyi AI Glasses, which attracted significant public interest [8][14]

Group 1: New AI Chip Releases - Suiyuan Technology and Muxi launched their new AI chips at the World Artificial Intelligence Conference (WAIC) on July 27, with Suiyuan's L600 chip developed over two and a half years, featuring a training and inference integrated architecture [1] - The L600 chip has a storage capacity of 144GB and a storage bandwidth of 3.6TB/s, supporting FP8 low precision to enhance training speed and reduce computing costs [1] - Muxi's new C600 chip also follows an integrated training and inference model, with a storage capacity increased from 64GB in the previous C500 to 144GB, utilizing advanced HBM3e memory [4] Group 2: Company Performance and Market Position - Suiyuan's previous S60 chip has shipped 70,000 units, making it a leading player in the domestic AI chip market, with deployments across five major intelligent computing clusters in China [1] - Muxi's C500 chip accounted for 97.28% and 97.87% of its revenue in 2024 and the first three months of 2025, respectively, indicating strong market demand for its products [7] - Muxi has sold over 25,000 GPU products as of March this year, and its IPO process is currently in the inquiry stage after being accepted on July 19 [7] Group 3: Future Developments - Suiyuan plans to establish a 10,000-card inference cluster in Gansu, set to be operational by December 2024, utilizing 10,016 S60 computing cards [1] - Muxi's C600 project has a total investment of approximately 1.37 billion yuan, with small batch production expected in the fourth quarter of this year, and the next-generation C700 series is already in development with an investment of about 2.04 billion yuan [4]

Core Insights - The company Jumpshare Star officially launched its new generation foundational model Step 3 at the World Artificial Intelligence Conference, aiming to enhance model adaptability and computational efficiency through collaboration with nearly 10 chip and infrastructure manufacturers [1][6] - Jumpshare Star has established a deep strategic partnership with Shanghai State-owned Capital Investment Co., which will participate in the company's latest financing round, focusing on strengthening capital connections and ecological business cooperation [1] - The company aims to achieve a revenue target of 1 billion RMB this year, with significant contract revenue already secured in the first half of the year [3] Revenue and Business Model - Jumpshare Star's Vice President indicated that the company has already secured several million RMB in contract revenue in the first half of the year, with a strong performance in confirmed revenue and gross profit [3] - The company has identified two unhealthy business models: customized delivery projects and pure API sales, which are affected by price wars and lack scalability [3] - The current commercialization strategy focuses on both B2B and B2C directions, targeting high-volume terminals such as cars and smartphones, and exploring IoT and embodied AI for future growth [3] Technological Developments - The Step 3 model features a total parameter count of 321 billion and an activation parameter count of 38 billion, designed for efficient inference across various hardware platforms [5] - The model's architecture allows for significant performance improvements, achieving up to 300% inference efficiency on domestic chips compared to previous models, and over 70% throughput improvement on NVIDIA Hopper architecture [5][6] - Jumpshare Star has authorized multiple domestic chip companies for Step 3 and has completed chip adaptation, with initial successful runs on Huawei Ascend chips and others [6][8] Ecosystem and Partnerships - The company has launched the "Model Chip Ecological Innovation Alliance" with major domestic chip manufacturers to streamline the technology across chips, models, and platforms, facilitating efficient large model solutions for enterprises and developers [6][8] - Jumpshare Star has partnered with leading financial institutions to develop specialized financial products, indicating a strategic focus on deep collaborations with top partners in various sectors [3]

Investment Rating - The report maintains an investment rating of "Outperform the Market - A" [5] Core Insights - The 2025 World Artificial Intelligence Conference (WAIC) showcased significant advancements in AI technology, particularly in computing power, algorithms, and applications, indicating a robust growth trajectory for the industry [12][60] - Domestic manufacturers are progressively enhancing their chip performance from inference to training, emphasizing ecosystem support and application scenarios [13][60] - The report highlights the emergence of quantum computing as a disruptive innovation in the computing power industry, with potential for exponential growth in computational capabilities [38][39] Summary by Sections 1. Industry Overview - The WAIC 2025 featured over 800 companies and more than 3000 cutting-edge exhibits, marking the largest scale in its history [12] - The conference focused on AI technology frontiers, industry trends, and global governance practices [12] 2. Computing Power - Domestic chip manufacturers are evolving from inference-focused AI chips to high-performance training chips, showcasing various industry applications [13] - The report notes the importance of supernodes, integrated machines, and liquid cooling solutions as key trends in the future of domestic AI computing power [18][19] 3. Algorithms - The evolution of large models is highlighted, with a shift towards multi-modal capabilities and the integration of deep reasoning and thinking abilities [43][44] - The report emphasizes the growing importance of AI agents and physical AI, which aim to enable intelligent agents to understand and interact with the real world [43][44] 4. Applications - The report identifies a flourishing landscape for both B2B and B2C AI applications, with significant advancements in areas such as AI glasses, smart driving, and robotics [60] - C-end applications focus on AIGC and productivity tools, while B-end applications emphasize cost reduction and efficiency improvements across various sectors [60][61] 5. Quantum Computing - Quantum computing is positioned as a critical future technology, with companies showcasing solutions that leverage quantum mechanics for superior computational capabilities [38][39]

Core Viewpoint - The World Artificial Intelligence Conference (WAIC) highlights the importance of embodied intelligence and the underlying computing infrastructure, which includes chips, boards, servers, and computing clusters, as a critical topic for the AI industry [2][5]. Group 1: Computing Infrastructure Transformation - This year's WAIC showcased a shift from a focus on "parameter competition" to practical discussions on "fragmented computing resource coordination," "low power and low cost," and "vertical product integration" [5]. - The trend of "full-chain localization" in computing infrastructure is gaining attention, driven by global supply chain disruptions and the need for self-sufficiency in core technologies [7][8]. - Domestic computing infrastructure manufacturers are expanding localization efforts beyond individual chips to encompass architecture design, software and hardware ecosystems, and industry implementation [9]. Group 2: Notable Companies and Innovations - Muxi, a representative of this localization trend, showcased its latest GPU, the Xiyun C600, which features a self-developed architecture and is designed for cloud AI training and inference [10]. - The Xiyun C600 is equipped with HBM3e memory, significantly enhancing memory bandwidth for large model training and inference [12]. - The previous generation Xiyun C500 series has already achieved full-chain localization in its server solutions, including compilers and interconnect protocols [13]. Group 3: High-Performance Computing Alternatives - Zhonghao Xinying presented its "Shan" series TPU, which utilizes controllable IP cores and a self-developed instruction set, achieving a 30% reduction in energy consumption for the same AI computing tasks [15]. - The "Taize" computing cluster system, based on the "Shan" TPU, can support over 400P (TF32) of floating-point operations, suitable for large AI model computations [17]. Group 4: Industry Applications and Collaborations - Huawei's "384 Super Node" was unveiled, showcasing its capability to interconnect 384 cards for large model adaptation, with over 80 models already developed [20]. - The collaboration between Huawei and partners spans various industries, including finance, healthcare, and transportation, demonstrating the practical applications of their computing solutions [20]. - Moore Threads presented 12 industry-specific demos, including a video super-resolution technology that enhances low-resolution video quality, showcasing the versatility of their solutions [21][22].

Core Insights - The World Artificial Intelligence Conference (WAIC) in Shanghai highlighted the growing importance of embodied intelligence, showcasing robots capable of physical interaction and complex tasks [1] - The focus has shifted from performance parameters to practical applications and solutions in the computing infrastructure sector, indicating a more mature industry approach [3][4] Group 1: Computing Infrastructure Trends - The computing infrastructure exhibition at WAIC displayed a significant change, moving away from the "parameter competition" seen in previous years [3] - Technical specifications are now integrated into industry solutions rather than being prominently displayed, reflecting a more application-oriented mindset [4] Group 2: Domestic Production and Self-Sufficiency - The push for "full-link domestic production" in computing infrastructure is gaining momentum, driven by global supply chain disruptions and technology challenges [5] - Domestic manufacturers are expanding their focus from individual chips to a comprehensive self-sufficient ecosystem, including architecture design and software integration [5] Group 3: Notable Companies and Innovations - Muxi, a representative company, showcased its latest GPU, the Xiyun C600, which features a self-developed architecture aimed at AI training and inference tasks [6] - The Xiyun C600 is equipped with HBM3e memory, enhancing bandwidth for large model training, although it was not yet available in server form at the exhibition [10] - Zhonghao Xinying presented its "Shan" TPU series, which boasts a fully controllable IP core and a design that reduces energy consumption by 30% for AI tasks [12] Group 4: Industry Applications and Collaborations - Huawei's "384 Super Node" technology was unveiled, showcasing its capability to support over 80 large models across various sectors, including finance and healthcare [18][19] - The collaboration between Huawei and partners aims to create solutions across multiple industries, enhancing the adaptability of their computing infrastructure [19] - Moore Threads demonstrated its video super-resolution technology, which significantly improves video quality and can be integrated into various applications, stimulating industry collaboration [20][23]

Core Insights - Yuntian Lifei plans to fully focus on AI chips by 2025, leveraging its expertise in neural network processor architecture and commercial chips [1] - The company aims to develop a domestic AI inference chip system characterized by high performance, low cost, and strong adaptability, targeting the AI inference market [1] Company Strategy - Yuntian Lifei will concentrate on three core areas: edge computing, cloud-based large model inference, and embodied intelligence [1] - The company is positioning itself as a leader in China's AI inference chip market, aiming to accelerate the deployment and rapid development of AI across various scenarios [1] Market Outlook - According to MarketsandMarkets, the global AI inference market is projected to reach approximately $106.15 billion (about 737 billion RMB) by 2025, with a compound annual growth rate (CAGR) of 19.2% from 2025 to 2030, potentially reaching around $254.98 billion by 2030 [6]

Group 1: AI Innovations and Applications - The 2025 World Artificial Intelligence Conference (WAIC) is themed "Intelligent Era, Common Ball" and showcases the latest AI technologies from various industry giants [2] - Guotai Junan launched the first full AI intelligent APP in the securities industry, named Guotai Junan Lingxi, which aims to enhance investment decision-making through automated content generation [4] - JD.com introduced the "Seven Fresh Kitchen" stores, utilizing AI robots for cooking to improve operational efficiency and service experience in the restaurant sector [2][4] Group 2: Transformation in the Restaurant Industry - The restaurant industry is undergoing significant changes due to high operational costs and efficiency challenges, prompting the need for innovative solutions [5] - The introduction of the "cooking robot" by Ecovacs' subsidiary, Tanke, addresses labor cost issues and enhances efficiency, allowing one person to operate five devices and reducing labor costs by 60% [5] - Tanke's cooking robot is seen as a revolutionary commercial model for the future of the restaurant industry, with a daily output exceeding 500 meals at its first experience store [5] Group 3: Development of Large Models - Shanghai Jueyue Star announced the release of its new foundational large model, Step-3, which will be open-sourced for global enterprises and developers [6] - The model aims to meet the needs of practical applications by focusing on strong intelligence, low cost, open-source capabilities, and multi-modal features [7] - The establishment of the "MoXin Ecological Innovation Alliance" by major players in the large model sector aims to enhance collaboration across the model, chip, and platform industries [6][8]

Core Viewpoint - The establishment of the "MoXin Ecological Innovation Alliance" by companies like Jiyue Xincheng, Huawei Ascend, and others marks a significant shift in China's AI and chip industry, moving from individual efforts to collaborative innovation in computing power [1][3][4] Group 1: Alliance Formation - The alliance signifies a transition from "individual combat" to "team collaboration" in the domestic computing power landscape, indicating a unified approach to tackle challenges in AI model development and chip compatibility [1][3] - The collaboration aims to integrate AI models with domestic chips, addressing the dependency on foreign technology and enhancing the overall ecosystem [3][5] Group 2: Key Players - Huawei Ascend is recognized as a leading player in the domestic chip market, successfully running the Step 3 model, showcasing its robust computing capabilities [4][5] - Other notable members like Muxi, Tianshu Zhixin, and Suiyuan Technology are acknowledged for their significant contributions to high-performance computing and AI training chips, representing the new force in domestic chip manufacturing [4][5] Group 3: Benefits of Collaboration - The alliance aims to standardize the adaptation of AI models across different chips, reducing the complexity for companies looking to utilize domestic computing power [5][6] - By aligning chip manufacturers and model developers, the alliance fosters a deep optimization of both hardware and software, enhancing performance and usability [5][6] Group 4: Economic and Security Implications - The collaboration is expected to lower the costs associated with AI services, making them more accessible to consumers and businesses alike [7] - Utilizing domestic chips and models enhances data security, particularly in sensitive sectors like finance and healthcare, by keeping data within national boundaries [7] Group 5: Future Outlook - While the alliance represents a significant step forward, challenges remain in terms of technology, ecosystem development, and market competition against international giants [8] - The initiative is seen as a foundational move towards establishing a self-sufficient and competitive domestic AI and chip industry, with long-term potential for growth and innovation [8]

Core Viewpoint - The rapid revenue growth of domestic AI model company Jiyue Xingchen, which is projected to reach nearly 1 billion in 2025, indicates a significant shift in the commercialization capabilities of domestic large models, potentially marking a new threshold for revenue in the industry [5][12][52]. Group 1: Event Overview - The WAIC World Artificial Intelligence Conference is a major annual event in the AI industry, attracting significant attention from both the industry and investment circles, with high-profile attendees including the Prime Minister [2]. - The conference serves as a platform for many important announcements from AI companies [4]. Group 2: Company Performance - Jiyue Xingchen, established only two years ago, has shown remarkable revenue growth, with last year's figures reportedly only in the tens of millions [7]. - The company focuses on foundational model research rather than easily monetizable AI applications, which makes its revenue growth even more impressive [8]. Group 3: Market Comparison - Other domestic large model companies have not disclosed revenue figures, but it is noted that Zhiyu's revenue was around 200-300 million last year, primarily from customized private sector services [9]. - The previous generation of AI companies typically took 3-4 years to surpass 1 billion in revenue, often leveraging rapid adoption in sectors like security and finance [11]. Group 4: Technological Advancements - Jiyue Xingchen has positioned itself as a leader in multi-modal capabilities, with its latest model, Step 3, achieving state-of-the-art results across various benchmarks [24]. - The Step series encompasses a wide range of functionalities, including text, speech, image, video generation, and more, making it versatile for numerous applications [25]. Group 5: Industry Collaboration - The company has formed a "Model and Chip Ecological Innovation Alliance" with nearly ten chip and infrastructure manufacturers, enhancing collaboration and technology sharing [35]. - This collaboration aims to optimize the performance of domestic chips in supporting large models, addressing the current gap between domestic and international chip capabilities [38]. Group 6: Market Applications - Jiyue Xingchen's technology is being integrated into various sectors, including automotive, mobile devices, and smart consumer electronics, with notable partnerships with major manufacturers [42][43]. - The demand for multi-modal capabilities is increasing, as evidenced by an 800% increase in model calls from smart terminals in the first half of the year compared to the previous half [48].