作为AI云原生平台，商汤大装置SenseCore致力于为用户提供敏捷、灵活、可靠的全栈AI基础设施服 务，推动大模型技术的高效落地与规模化应用。此次商汤大装置SenseCore与昇腾384超节点的成功适 配，让多租户、大规模、弹性AI云服务成为可能。未来，双方还将探索更多应用场景，包括大模型推 理加速、智能体应用部署、面向垂直行业的大模型训练与推理优化等，进一步加速基于SenseCore的昇 腾384超节点在各行各业的应用落地。 北京商报讯（记者 魏蔚）9月4日，商汤科技宣布，商汤大装置SenseCore与昇腾384超节点完成全面适 配。超节点（SuperPod）是一种通过高速互联技术，将多个GPU（图形处理器）/NPU（嵌入式神经网 络处理器）整合为统一计算单元的新型架构，解决AI大模型训练中的算力协同与通信效率问题。昇腾 384超节点（Atlas 900 A3 SuperPoD）是华为推出的业界最大规模超节点方案，这一全新方案架构，对 软件栈的升级和平台调度优化提出了更高要求。 ...

Core Insights - The 2025 World Artificial Intelligence Conference (WAIC) in Shanghai showcased the growing interest in the AI industry from academia, industry, and investors, reflecting a significant shift in focus since the first conference in 2018 [3][4] - The AI landscape is evolving with a shift from homogeneous models to differentiated products, as companies face increasing competition and seek to maintain core advantages [6][7] - Traditional internet giants like Alibaba and Tencent are promoting full-stack AI capabilities, integrating AI models into comprehensive solutions to lower barriers for enterprises [10] Group 1: AI Model Development - The competition among large model companies has intensified, leading to a focus on differentiated product strategies, such as long text processing and multi-modal capabilities [6] - The boundaries between "model" and "application" are blurring, with model companies transitioning to comprehensive AI platforms that integrate various technologies [7] - The emergence of open-source models like DeepSeek R1 is reshaping the competitive landscape, prompting companies to explore new business models [6][7] Group 2: Cloud and AI Integration - Established cloud providers are building AI ecosystems around their models, offering one-stop AI solutions that enhance the accessibility of AI capabilities for enterprises [10] - The shift towards public cloud platforms for AI capabilities is driven by the challenges of private cloud deployment, particularly for small and medium-sized enterprises [10] - The integration of AI modules into existing cloud infrastructures is expected to reshape asset valuation logic in the long term [10] Group 3: Domestic Computing Power - Domestic computing power was prominently featured at the conference, with Huawei showcasing its Ascend 384 super node, which boasts double the computing power of NVIDIA's GB200 NVL72 system [13] - Domestic GPUs are increasingly competitive, although challenges remain in memory bandwidth and interconnect capabilities [14] - The demand for private AI deployment is driving innovation in AI integrated machines, reflecting a strong market need [15] Group 4: Edge AI Development - The WAIC highlighted the innovative potential of edge AI, though the commercialization paths for these products remain limited [18] - Key areas for improvement in edge AI include multi-modal perception and decision-making capabilities, which are critical for applications like robotics and AR [18] - The smartphone is positioned as a likely platform for AI agents due to its proximity to users and strong computational capabilities, although industry caution regarding technology maturity persists [18][19]

Core Insights - The year 2025 is anticipated to be a pivotal year for the deployment of large AI models, with integrated machines serving as essential infrastructure for businesses to implement these models effectively [1][2] - The demand for AI integrated machines is driven by both market needs and technological advancements, providing a comprehensive solution that combines hardware, software, and algorithms [1][2] Industry Trends - AI integrated machines are designed to facilitate the application and deployment of large AI models, offering lower costs and easier operation compared to traditional server clusters, making them suitable for small and medium enterprises [2][5] - The trend indicates that AI integrated machines will evolve towards higher efficiency, lower energy consumption, and easier deployment, with specialized products tailored to various industry needs [1][2] Product Developments - The launch of the Langxin Jiugong AI Energy Integrated Machine aims to address the challenges in the energy sector, providing a closed-loop solution that enhances AI application in power trading and billing management [2][3] - The Langxin Jiugong AI Energy Integrated Machine features nine core functions, including precise load forecasting and intelligent energy scheduling, and is fully domestically produced to ensure safety and autonomy in energy applications [3][4] Market Potential - According to research, 23% of central enterprises have already deployed large models, with expectations for further growth in adoption rates [5] - The demand for integrated machines is projected to reach 150,000 units in 2025, 390,000 units in 2026, and 720,000 units in 2027, with a market potential for central state-owned enterprises estimated at 123.6 billion yuan, 293.7 billion yuan, and 520.8 billion yuan respectively [5] National Strategy - The push for a fully domestic AI industry chain is seen as essential for national security, with a focus on developing independent and controllable AI systems across various sectors [6][8] - The Shanghai government has introduced measures to support AI application expansion, including financial incentives for computing power usage and model deployment, to address the growing demand for AI capabilities [7][8]

Core Insights - The year 2025 is anticipated to be a pivotal year for the deployment of large AI models, with integrated machines serving as a crucial component for enterprises to implement these models effectively [1][2] - The demand for AI integrated machines is driven by both market needs and technological advancements, providing a complete solution that lowers barriers for enterprises to deploy applications quickly [1][2] Group 1: AI Integrated Machines - AI integrated machines are designed specifically for the application and deployment of large AI models, combining hardware and software components, making them more cost-effective and user-friendly for small to medium-sized enterprises [2][4] - The launch of the Langxin Jiugong AI Energy Integrated Machine aims to address the challenges faced by the energy sector, such as lack of computing power and complex model deployment, by providing a high-integrated, stable, and user-friendly solution [2][3] - The Langxin Jiugong AI Energy Integrated Machine features nine core functions, including precise load forecasting and intelligent energy scheduling, and is fully domestically produced to ensure safety and autonomy in energy applications [2][3] Group 2: Market Trends and Projections - As of now, 23% of central enterprises have deployed large models, with expectations for further increases in adoption rates, particularly driven by the DeepSeek model [4] - Demand for integrated machines is projected to reach 150,000 units in 2025, 390,000 units in 2026, and 720,000 units in 2027, with a market potential for central state-owned enterprises estimated at 123.6 billion yuan, 293.7 billion yuan, and 520.8 billion yuan respectively [4] Group 3: Domestic Production and Supply Chain - The push for a fully domestic AI industry chain is seen as essential for national security, with a focus on ensuring that AI models and technologies are under local control [5][6] - The Shanghai municipal government has introduced measures to support AI applications, including financial incentives for computing power usage and model deployment, to address the supply-demand imbalance in AI computing resources [6][7] Group 4: Challenges and Strategic Directions - The current core contradiction in China's AI computing landscape lies between the rapidly growing demand for model training and the limited supply of high-end computing resources [7][8] - Industry experts emphasize the need for a dual approach in AI computing development, focusing on both independent innovation in core technologies and maintaining open collaboration for international standards [8]

Core Insights - The year 2025 is anticipated to be a pivotal year for the deployment of large AI models, with integrated machines serving as a crucial component for enterprises to implement these applications effectively [1][2] - The demand for AI integrated machines is driven by both market needs and technological advancements, providing a complete solution that lowers the barriers for enterprises to adopt AI [1][2] Industry Trends - AI integrated machines are designed to facilitate the application and deployment of large AI models, combining hardware, software, and algorithms into a single device, making them more accessible for small and medium enterprises [2][4] - The trend indicates that AI integrated machines will evolve towards higher efficiency, lower energy consumption, and easier deployment, with specialized products tailored to various industry needs [1][2] Market Demand - The energy sector is experiencing a surge in demand for AI applications, with integrated machines like the Langxin Jiugong AI Energy Integrated Machine being introduced to address challenges such as computational power shortages and complex model deployments [2][3] - The deployment of AI integrated machines is expected to grow significantly, with projected demand reaching 150,000 units in 2025, 390,000 units in 2026, and 720,000 units in 2027, indicating a substantial market opportunity [4] Technological Advancements - The Langxin Jiugong AI Energy Integrated Machine features a self-developed AI model that integrates time-series prediction and AI agent technologies, offering nine core functionalities for optimizing energy management [2][3] - The introduction of integrated machines is seen as a solution to the challenges of computational power supply and the need for a reliable, controllable technology base in critical sectors like energy [3][5] Policy Support - Local governments are increasing support for AI applications, with initiatives such as issuing computing power vouchers and subsidies for renting intelligent computing resources to stimulate the AI ecosystem [6][7] - The Shanghai Municipal Economic and Information Commission has introduced measures to lower the cost of intelligent computing and support the development of AI models, indicating a strong governmental push towards AI integration [6][7] National Strategy - The push for a fully domestic AI industry chain is viewed as essential for national security, with a focus on developing self-sufficient technologies in critical areas such as chips and algorithms [5][8] - The industry is encouraged to balance core technology innovation with open collaboration to avoid dependency on foreign technologies while participating in international standards [8]

Core Viewpoint - Huawei's Ascend 384 SuperNode is a significant advancement in AI computing, showcasing the shift from single-chip performance to large-scale system efficiency and ecosystem integration [1][2]. Group 1: Huawei's Ascend SuperNode - Huawei showcased the Ascend 384 SuperNode at the 2025 World Artificial Intelligence Conference, which features a high-bandwidth, low-latency interconnection of 384 NPUs [1]. - The Ascend SuperNode is recognized as the largest in the industry and has been commercially deployed in data centers like China Telecom, indicating a shift in global AI computing capabilities [2]. Group 2: Market Impact - The expansion of Huawei's Ascend supply chain is expected to benefit sectors such as optical communication, high-speed connection modules, domestic computing, AI applications, and data centers [2].

Core Insights - The 2025 World Artificial Intelligence Conference (WAIC 2025) concluded successfully in Shanghai, featuring over 1,500 top experts from more than 70 countries and regions, showcasing over 3,000 cutting-edge technologies, with more than 100 global and Chinese product debuts [2] - The conference highlighted a shift in the AI industry from "parameter competition" to "value creation," indicating China's AI sector is moving towards defining global standards and promoting inclusive development [2] - Huawei showcased its Ascend 384 SuperNode (Atlas 900 A3 SuperPoD), which addresses communication bottlenecks in AI model training and inference through high bandwidth and low latency interconnects among 384 NPUs [2] AI Model Development - Over 80 large models have been developed based on Huawei's Ascend platform, with partnerships involving over 2,700 collaborators to create more than 6,000 industry solutions across 11 sectors, including internet, finance, and healthcare [4] - Zhaoxin presented its KX-7000N AIPC processor, which integrates high-performance NPUs to enhance local AI processing capabilities for applications like speech recognition and image generation [4][6] AI Hardware Innovations - Moore Threads showcased its full-featured GPU solutions, emphasizing their role in general-purpose accelerated computing platforms and various industry applications, including professional graphics rendering and AI model training [6][8] - Muxi introduced its flagship GPU, the Xiyun C600, which is designed to provide reliable computing power for critical sectors, marking a significant step in domestic supply chain development [8][10] Industrial AI Applications - Siemens made its debut at WAIC, presenting its Industrial Copilot system, which integrates large language models with industry knowledge to enhance production efficiency by 50% [11][15] - The company also demonstrated AI applications in energy systems and process industries, showcasing its extensive patent portfolio and AI expertise [15] Ecosystem Development - North Electric Intelligence introduced a comprehensive AI empowerment system that integrates city-level AI infrastructure with industry-specific solutions, aiming to enhance digital transformation across various sectors [11][13] - Super Fusion launched its FusionOne AI solution, which aims to bridge the gap in AI implementation across industries by providing a full-stack digital resource intelligent system [15][17] Consumer AI Products - Rokid unveiled its lightweight AR glasses, which utilize advanced technologies for seamless user interaction, showcasing the potential of AI in enhancing user experiences in everyday scenarios [19]

Core Insights - The 2025 World Artificial Intelligence Conference (WAIC) showcased a shift in focus from pure technical parameters to practical applications and commercial value in AI technology [2][14] - The competition in computing power is evolving into a comprehensive system engineering challenge, addressing performance, compatibility, storage, and energy efficiency [4][10] - AI companies are increasingly integrating their models with real-world applications to unlock new data sources and enhance AI capabilities [15][16] Computing Power Infrastructure - Companies like Huawei and China Digital are pushing the limits of computing power, with Huawei's Atlas 900 A3 SuperPoD achieving a performance of 300 PFLOPS [2][4] - The financial sector is supporting AI infrastructure, with companies like Chip Xin Leasing investing 8 billion yuan in AI-related projects [4] - The demand for private deployment of large models is increasing due to data security concerns, indicating a shift in market needs [5][6] Model and Application Development - AI model developers are focusing on deep integration with industry scenarios to create real business value, moving away from mere technical showcases [14][17] - Companies like Step Leap Star are launching new models aimed at cost reduction and efficiency improvement, collaborating with multiple chip manufacturers to enhance compatibility [17][18] - The importance of data storage and management is highlighted, with companies like Dawning Storage addressing challenges in data accessibility and efficiency [8][9] AI in Creative Industries - AI-generated content (AIGC) is transforming creative processes, with companies like Digital Kingdom introducing platforms that streamline content creation [20][21] - AI is positioned as a "super assistant" for creators, enhancing productivity while allowing them to focus on core creative tasks [21] Consumer-Focused AI Products - New AI products, such as the TicNote AI recording pen, are being developed to serve individual users, encapsulating complex AI capabilities in user-friendly formats [23] - The overarching goal of AI advancements is to contribute to real GDP growth across society, industries, and nations [24]

Group 1: AI Industry Trends - The 2025 World Artificial Intelligence Conference (WAIC) showcased over 800 companies and highlighted a growing competition in computing power infrastructure led by hardware manufacturers like Huawei and Inspur [2][3] - There is a shift in focus from pure technical specifications to the practical application of AI in business scenarios, emphasizing the importance of translating AI technology into real commercial value [2][10] - Companies are increasingly investing in AI infrastructure, with Chip Xin Leasing announcing an investment of 8 billion yuan to support data centers and computing power leasing [3] Group 2: Computing Power and Infrastructure - Huawei introduced the Ascend 384 SuperNode (Atlas 900 A3 SuperPoD) with a computing power of 300 PFLOPS, representing a significant advancement in computing capabilities [3][4] - The competition in computing power is evolving into a complex system engineering challenge that requires addressing performance, compatibility, storage, and energy efficiency [3][4] - The demand for private deployment of large models is increasing due to data security concerns, leading to differentiated computing power solutions for training and inference [5] Group 3: Data Storage and Management - Data centers face challenges such as data silos and inefficient data utilization, prompting companies like Inspur to develop unified storage systems to enhance data accessibility [6][7] - The collaboration between China Mobile and Inspur aims to establish a nationwide unified file storage system to support the "East Data West Computing" strategy [6] - Efficient data management is crucial for maximizing computing power, with investments in storage systems yielding significant cost savings in computing expenses [7] Group 4: AI Model Development and Application - AI model developers are moving away from mere parameter comparisons to deeply integrating AI into industry applications, seeking to create ecosystems that generate commercial value [10][11] - Companies like Step Leap Star are focusing on cost reduction and efficiency improvement in AI models, collaborating with multiple chip manufacturers to enhance hardware compatibility [11] - The integration of AI into creative industries is transforming production processes, with companies like Digital Kingdom launching platforms to streamline content creation [14] Group 5: Consumer Applications of AI - AI technologies are being packaged into consumer-friendly products, such as the TicNote AI recording pen, which aims to assist users in various tasks [16] - The overarching goal of AI advancements is to contribute to real GDP growth across society, industries, and nations, rather than merely showcasing intelligence [16]

Group 1 - The 2025 World Artificial Intelligence Conference (WAIC 2025) commenced in Shanghai, attracting over 1,572 global leaders from 73 countries and regions, including 12 top award winners and over 80 leading academicians [1] - The event features a themed street area called "WAIC 2025 Alley," showcasing various robots that captivated a large audience [6] - The exhibition area of this year's conference exceeded 70,000 square meters, with over 800 companies showcasing more than 3,000 cutting-edge exhibits, including over 40 large models and 60 smart robots, marking the largest scale in the event's history [56] Group 2 - The conference's core sub-venue, Xuhui West Bank, focused on the construction of an artificial intelligence innovation ecosystem and hosted a high-level meeting on global AI governance [8] - The event highlighted the rise of domestic chip manufacturers, attracting increasing attention from both domestic and international exhibitors [30] - Huawei showcased its Ascend 384 SuperNode, which utilizes a supernode architecture to achieve high-bandwidth, low-latency interconnection among 384 NPUs, addressing communication bottlenecks within clusters [36]