2025.09.29 本文字数：2124，阅读时长大约4分钟 作者 |第一财经 李娜 尽管资本市场对算力泡沫发出的警告不断，但英伟达CEO黄仁勋依然对算力产业的发展调高了预期。在 近期一次访谈中，他预测AI推理即将迎来十亿倍的增长，并指出由AI增强的经济活动将催生一个每年5 万亿美元的AI基础设施资本支出市场。 值得注意的，就在访谈前的72小时，英伟达与OpenAI正式达成战略合作，黄仁勋称这笔交易为"能想象 到的最聪明的投资之一"。但也有券商分析师将这种交易模式比喻为金庸小说中的"梯云纵"——看似向 OpenAI投资1000亿美元，实则这些资金又以订单形式流回自家口袋。这种左脚踩右脚的模式，使英伟 达提前锁定利润。 面对这样的"算力激进者"，比起泡沫更担心的是错过，近一个月以来，如甲骨文、英特尔等厂商开始以 订单、资金或股权方式与英伟达结盟，一个以美国企业主导的AI新生态正在酝酿。而在国内算力市场 上，随着单颗算力芯片的焦虑逐步缓解，建立一个具有同样生命力和吸引力的生态系统也开始变得迫 切。 生态竞争进入拐点 不断被抬高的需求正在驱动全球算力竞赛进入淘汰赛阶段。 在英伟达与OpenAI的合作中，计划建设高达 ...

对于国内算力市场，建立自己的生态系统也开始变得迫切。 尽管资本市场对算力泡沫发出的警告不断，但英伟达CEO黄仁勋依然对算力产业的发展调高了预期。在 近期一次访谈中，他预测AI推理即将迎来十亿倍的增长，并指出由AI增强的经济活动将催生一个每年5 万亿美元的AI基础设施资本支出市场。 值得注意的，就在访谈前的72小时，英伟达与OpenAI正式达成战略合作，黄仁勋称这笔交易为"能想象 到的最聪明的投资之一"。但也有券商分析师将这种交易模式比喻为金庸小说中的"梯云纵"——看似向 OpenAI投资1000亿美元，实则这些资金又以订单形式流回自家口袋。这种左脚踩右脚的模式，使英伟 达提前锁定利润。 面对这样的"算力激进者"，比起泡沫更担心的是错过，近一个月以来，如甲骨文、英特尔等厂商开始以 订单、资金或股权方式与英伟达结盟，一个以美国企业主导的AI新生态正在酝酿。而在国内算力市场 上，随着单颗算力芯片的焦虑逐步缓解，建立一个具有同样生命力和吸引力的生态系统也开始变得迫 切。 生态竞争进入拐点 不断被抬高的需求正在驱动全球算力竞赛进入淘汰赛阶段。 这一决策也意味着华为主动斩断了一条可能的"利润路径"，为了获得互联网等企业的 ...

Core Viewpoint - The article discusses the challenges and opportunities in China's computing power market, particularly focusing on the emergence of "intelligent computing centers" and the role of the company Wuwen Xinqun in addressing these challenges through innovative solutions [1][10]. Group 1: Current State of Computing Power - As of September 2024, China's computing power scale reached 246 EFLOPS, with intelligent computing power growing over 65% year-on-year, and over 13,000 computing power application projects across various industries [1]. - Despite the growth in computing power, the average cabinet utilization rate in intelligent computing centers is only 20% to 30%, with some enterprise-level centers as low as 10% [1][2]. - The overall utilization rate of computing power in the country is only 32%, indicating a significant gap between supply and demand [2]. Group 2: Challenges in the Computing Power Market - There is a shortage of quality computing power supply, making it difficult for many enterprises to find suitable resources [2][3]. - High usage thresholds prevent startups from effectively utilizing available computing power, leading to a phenomenon of "computing power scarcity" among AI companies [3]. - The domestic chip ecosystem is fragmented, with various models and architectures that are incompatible, hindering efficient resource flow [3]. Group 3: Wuwen Xinqun's Approach - Founded in May 2023 by a team with strong ties to Tsinghua University, Wuwen Xinqun has quickly gained market attention, securing nearly 1 billion yuan in funding within two years [4][5]. - The company aims to be a "computing power operator" in the era of large models, addressing the challenges posed by the dominance of NVIDIA's CUDA ecosystem and the fragmentation of domestic chip manufacturers [5]. - Wuwen Xinqun has developed a cloud computing network that integrates heterogeneous computing resources, allowing developers to focus on applications without worrying about underlying hardware differences [5][10]. Group 4: Product Offerings - Wuwen Xinqun has introduced three core products, referred to as "three boxes," to enhance computing power utilization across various scales [6][9]. - The "Wuqiong AI Cloud" serves as a large-scale computing network, integrating resources from 26 provinces and over 53 data centers, with a total computing power exceeding 25,000 P [7]. - The "Wujie Intelligent Computing Platform" targets large computing clusters, demonstrating significant performance in training large models [8]. - The "Wuyin Terminal Intelligence" solution is designed for limited computing terminals, enabling them to perform complex tasks without relying on cloud resources [9]. Group 5: Future Outlook - Wuwen Xinqun's efforts reflect a broader need for a cohesive ecosystem that allows computing power to flow effectively, addressing the current challenges of fragmentation and high costs [10][11]. - The company aims to enhance the actual utilization rate of computing resources and improve the cost-performance ratio, contributing to the growth of China's AI industry [10].

Core Viewpoint - The release of DeepSeek's V3.1 model highlights a significant shift in the domestic AI ecosystem, emphasizing the collaboration between software and hardware through the introduction of the UE8M0 FP8 floating-point format designed for next-generation domestic chips [6][11]. Group 1: Technical Innovations - DeepSeek's V3.1 model features a mixed reasoning architecture, improved thinking efficiency, and enhanced agent capabilities [6]. - The UE8M0 FP8 format prioritizes numerical range stability over decimal precision, allowing for stable training on domestic GPUs, which differ from NVIDIA's hardware [11][13]. Group 2: Competitive Landscape - The competition around FP8 standards represents a broader struggle for computational ecosystem dominance, with NVIDIA solidifying its position through the Blackwell architecture and MXFP8 format [13]. - Domestic AI firms are encouraged to innovate collaboratively from models to chips, as seen with companies like Muxi and Suiyuan Technology optimizing FP8 computation efficiency [13][18]. Group 3: Future Trends - The collaboration between hardware and software, even at the cost of temporary performance compromises, is likely to become a prevailing trend in the industry [14]. - The UE8M0 FP8 format signifies a critical step for the domestic AI industry, moving from isolated breakthroughs to comprehensive stack collaboration, which may offer more long-term value than merely increasing model scale [16].

Core Viewpoint - The event focused on the deep integration of intelligent computing technology and industry, highlighting the launch of new products by Huawei and Hengyang Data to empower various sectors through advanced computing solutions [1][3]. Group 1: Event Overview - The 2025 Huawei Shenzhen Government and Enterprise and Hengyang Data New Product Launch Conference successfully took place in Shenzhen, centered around the theme "Kunpeng Ascending, Empowering Business Renewal" [1]. - The event attracted numerous industry elites to discuss how intelligent computing can inject new vitality into various sectors [1]. Group 2: Product Launch - Huawei and Hengyang Data jointly launched two significant products: the Hengyang Data K+A integrated machine and the SempFusion intelligent computing platform, marking a new stage in their collaboration in the intelligent computing field [1][6]. - The K+A integrated machine is the first flagship model officially released by a partner in South China, designed to assist enterprises in rapidly achieving intelligent transformation and application implementation [4]. Group 3: Strategic Collaboration - Huawei's commitment to building an open and collaborative computing ecosystem through its core technologies, Kunpeng and Ascend, was emphasized by the Minister of Huawei's Commercial Marketing and Solutions Sales Department [3]. - Hengyang Data's Chairman highlighted the joint development of a series of intelligent computing products and solutions aimed at empowering enterprises' digital transformation [3]. Group 4: Technological Advancements - Huawei's marketing expert discussed the creation of competitive differentiated products in collaboration with numerous KPN partners, covering various application scenarios [3]. - The introduction of large models like DeepSeek has significantly lowered the barriers to AI deployment and usage, accelerating the pace of industry intelligence [3]. Group 5: Ecosystem Development - The collaboration aims to create an open and collaborative computing ecosystem, consolidating the efforts of the entire industry chain to meet diverse industry demands [1][3]. - The SempFusion intelligent computing platform is designed to provide enterprises with a full-stack AI solution, enhancing knowledge assets and empowering business activities [6]. Group 6: Practical Applications - Representatives from Eden Software and Mingtai Intelligent shared their practical achievements in the intelligent computing field, showcasing products that provide end-to-end smart office capabilities and AI applications in government scenarios [6]. - Attendees experienced a series of high-performance multi-core integrated computing platforms and related software products developed by Hengyang Data based on Huawei's computing foundation [6].

Core Insights - The event themed "Kunpeng Ascending Everywhere, Achieving Great Business" was successfully held, focusing on the integration of intelligent computing technology with industry [1] - Huawei and Hengyang Data jointly launched the Hengyang Data K+A integrated machine and SempFusion intelligent computing platform [1][8] Industry Trends - Global computing power demand is increasing exponentially, necessitating a shift from single technology breakthroughs to deep empowerment across various industry scenarios [2] - The need for ecological collaboration to break down "technology silos" and foster innovation in the industry is emphasized [2] Company Strategies - Huawei aims to build an open and collaborative computing ecosystem through its core technologies, Kunpeng and Ascend, to enhance the usability of computing power [4] - Hengyang Data's collaboration with Huawei focuses on creating innovative products and solutions in the intelligent computing field, facilitating digital transformation for enterprises [6] Product Launches - The K+A integrated machine, a flagship model combining SK90 intelligent computing machine and Ascend acceleration card, was officially launched, marking a significant step in enhancing computing power usability [8] - The SempFusion intelligent computing platform was unveiled, representing a new phase of collaboration between Huawei and Hengyang Data, integrating hardware and software solutions [8] Ecosystem Development - The partnership between Huawei and Hengyang Data demonstrates the commercial value of their technologies and the power of a collaborative ecosystem [9] - Various companies showcased applications of AI in different sectors, leveraging Huawei's computing power foundation to enhance business capabilities [9]

Group 1 - The meeting focused on optimizing industrial functional zones in Changsha to enhance industrial agglomeration, based on a proposal by a local representative [1][3] - The current phase of industrial development in Changsha is characterized by structural optimization, new and old kinetic energy conversion, and high-quality development, while also recognizing existing shortcomings [3][4] - Emphasis on planning and systematic thinking to support project development, targeted investment, and optimizing the industrial ecosystem [3][4] Group 2 - The importance of high-quality handling of representative suggestions to ensure effective implementation and transformation into tangible development outcomes [4] - Research conducted on the development of the new generation computing system industry chain in Changsha, highlighting the need for departmental collaboration and technological synergy [6][11] - Encouragement for local enterprises to focus on their strengths and invest in core technology development to gain competitive advantages in the industry [11]

Core Insights - The company, Wunwen Xinqiong, launched a comprehensive AI efficiency enhancement solution during the WAIC 2025, introducing three core products: Wunqiong AI Cloud, Wujie Intelligent Computing Platform, and Wuyin Terminal Intelligence [2][3] - The Wunqiong AI Cloud targets global computing networks ranging from 10,000 to 100,000 cards, while the Wujie Intelligent Computing Platform is designed for large computing clusters of 100 to 1,000 cards, and the Wuyin Terminal Intelligence focuses on limited computing terminals from 1 to 10 cards [2][3] - The company aims to create a "universal language" for the industry to enable seamless communication and collaboration between different chip architectures, thereby enhancing the efficiency of AI technology deployment [3][4] Product Offerings - Wunwen Xinqiong's product ecosystem spans from infrastructure to industry applications, supporting AGI technology's large-scale implementation [6] - The company reported that its AI cloud platform provides a one-stop service for enterprises and developers, including cloud management, foundational cloud products, and large model development platforms [5][6] - The company has achieved significant performance optimization for over ten types of domestic AI chips, enhancing their performance by 50% to 200% through algorithm and compilation optimizations [5][6] Market Position and Growth - Wunwen Xinqiong completed nearly 500 million yuan in Series A financing in 2024, setting a record for the largest single financing in domestic AI infrastructure [5] - The company has raised over 1 billion yuan since its establishment in May 2023, indicating strong investor confidence and market potential [5] - The company serves the world's largest AI incubator, Shanghai Mosu Space, which has surpassed 10 billion daily token calls, supporting over 100 innovative AI applications [6] Strategic Vision - The CEO emphasized the importance of creating a closed-loop ecosystem for chip and model development to guide future advancements in the industry [4][8] - The company aims to achieve its ultimate vision of "limitless intelligence and precise computing" by harmonizing scene scale, computing resources, and intelligent efficiency [9]

Group 1: Beverage Industry Insights - The beverage industry is expected to experience a peak season with concentrated catalysts in the short term, alongside accelerated product launches throughout the year, aligning with the trends of health consciousness and value-for-money [3] - Recommended investment strategies include focusing on high-quality leaders such as Nongfu Spring, Dongpeng Beverage, and Uni-President China, as well as targets with high safety margins and improvement potential like China Resources Beverage, Liziyuan, Chengde Lulule, Huanyoujia, Junyao Health, and Chongqing Beer [3] Group 2: Imaging Equipment Sector - The report highlights Ying Shi Innovation (688775.SH) as a leader in smart imaging devices, emphasizing its focus on panoramic and action cameras, with a robust technological framework including stabilization and AI image processing [4] - The company's revenue grew from 1.328 billion in 2021 to 5.574 billion in 2024, with a CAGR of 61.31%, while its net profit increased from 252 million to 946 million during the same period, achieving a CAGR of 55.42% [4] Group 3: Automotive Sector - Yadi Holdings (01585.HK) is positioned to benefit from policy-driven demand recovery, with expectations for a return to rapid growth in the industry due to trade-in programs and new national standards [7] - The company is projected to achieve net profits of 2.93 billion, 3.55 billion, and 4.11 billion from 2025 to 2027, with corresponding PE ratios of 12.1, 10.0, and 8.6, respectively, and has been rated as a "buy" [7] Group 4: Computing Sector - The report discusses the strategic merger between Haiguang Information and Zhongke Shuguang, which is expected to reshape the domestic computing power ecosystem, with a focus on the server sector [8] - Key players to watch in the computing power ecosystem include Zhongke Shuguang, Haiguang Information, Inspur Information, and others [8]

Core Viewpoint - The article discusses the recent U.S. export restrictions on semiconductor technology to China, particularly focusing on the impact on the electronic design automation (EDA) market and the competitive landscape in the computing power sector between the U.S. and China [2][8]. Group 1: U.S. Export Restrictions - The U.S. government has cut off certain semiconductor design software exports to China, affecting major EDA companies like Cadence, Synopsys, and Siemens, which hold over 80% of the Chinese EDA market [2]. - The U.S. has implemented stricter AI chip export controls, categorizing China under a comprehensive ban on GPU chips [8]. Group 2: Domestic Developments in China - Domestic companies are actively competing in the computing power ecosystem, with significant developments such as the merger between Haiguang Information and Zhongke Shuguang, both of which are industry leaders [5][21]. - The merger aims to integrate chip design and server manufacturing, enhancing the domestic computing power ecosystem and addressing supply chain security concerns [24][25]. Group 3: Market Dynamics and Competition - NVIDIA's market share in China has dropped from 95% to 50% due to U.S. export controls, allowing local companies like Huawei and Cambricon to gain ground [10][13]. - Despite the loss in market share, there remains a significant dependency on NVIDIA's software ecosystem, particularly the CUDA platform, which is widely used in AI model development [15][18]. Group 4: Strategic Implications - The merger of Haiguang and Shuguang represents a strategic move to create a comprehensive solution that combines chip design, server manufacturing, and cloud services, similar to Huawei's approach with its Ascend chips and cloud services [26][28]. - China's strategy contrasts with the U.S. approach, focusing on building a self-sufficient computing power ecosystem while ensuring that domestic industries utilize local products [31][32]. Group 5: Future Outlook - The article suggests that while China is currently in a defensive position in the computing power competition, the integration of companies like Haiguang and Shuguang is a step towards strengthening its industrial capabilities [34][35]. - The focus will need to shift towards developing a robust software ecosystem to complement the hardware advancements, as the U.S. still holds a significant advantage in this area [36][38].