今年7、8月，我国月度全社会用电量连续突破万亿千瓦时大关。其中，在移动互联网、大数据、云计算 等带动下，互联网和相关服务业用电量同比增加。此外，由于新能源汽车的渗透率不断提升，充换电服 务业用电量也同比显著增长。 无论是算力还是新能源汽车，都展现出对电力的巨大需求。那么，在这一趋势下，我国的电力系统能否 扛住冲击？ 一、制造业驱动电力需求增长 在接受21世纪经济报道能源策工作室采访时，厦门大学教授、中国能源政策研究院院长林伯强认为，中 国月度社会用电量突破一万亿千瓦时是必然事件。 "它（社会用电量）本来跟经济发展，跟我们制造业是息息相关的。与此同时，居民生活水平不断提 高，用电量也会提高。"林伯强表示，当前我国电力需求结构中，来自工业、制造业的用电需求特别 大。"目前数据中心肯定是带来用电影响，但目前也没有那么大。" （原标题：五年之内，算力会对电力造成冲击？| 21新能说） 21世纪经济报道记者曹恩惠 全球算力正以指数级速度增长。 根据行业数据，当前全球数据中心的年耗电量已占全球总用电量的3%以上。而随着AI大模型的快速发 展，这一比例还将继续上升。 当全球算力需求快速增长之时，电力供应的稳定性与经济性，成 ...

第一：七部门发布，算力、人工智能再度迎来利好 昨天，国家7部门联合发布，深入推动服务型制造创新发展方案，提出要深化"5G+工业互联网"融合应用，按需布局算力基础 设施，提升工业数据要素供给，推动人工智能技术与服务型制造融合创新， 势必未来在算力、人工智能上面的投入不会少，对高科技板块都是利好。不过美高科技大跌，能抵消影响吗？ 第四：周末愉快 昨天是正常上班，今天小放一天假期，感觉工作节奏被打乱了不少。 不过，不要紧，反正都是享受人生的日子，努力是一天，不努力也是一天。为何不选择全力以赴呢？ 我是股民聪哥，一个立志花10年让1亿粉丝轻松看懂财经的男人，关注我，一起向上成长。 以上仅为个人看法，不作为任何建议！ 由于预算"谈不拢"，美联邦ZF持续"停摆"。有工作人员表示，已经正式启动了裁员程序，如果美时间周一还没有化解危机， 15日起将暂停军队薪资。 停摆不是历史上第一次，当然也不是最后一次。两个"人"经常闹意见不统一，正常得很。只是，对经济伤害太大，美联储下 次降息几乎板上钉钉。 第三：刚果（金）钴出口禁令将于下周解除 根据周六的最新消息，刚果（金）宣布，持续了8个月的钴出口禁令将于下周（10月16日）解除。 ...

Group 1: OpenAI Developments - OpenAI's video AI application Sora achieved over 1 million downloads within 5 days of launch, surpassing the download speed of ChatGPT [1] - Sora is currently limited to iOS devices and operates on an invitation-only basis, which contrasts with ChatGPT's initial public release [1] - OpenAI is enhancing ChatGPT into a "super app" that allows users to directly access third-party applications within conversations [6] Group 2: Strategic Partnerships and Investments - AMD has issued a warrant to OpenAI allowing the purchase of up to 160 million shares at $0.01 each, as part of a strategic partnership to deploy 6 gigawatts of AMD GPU capacity [2] - NVIDIA is also collaborating with OpenAI, with plans to invest up to $100 billion, contingent on the construction of data centers [2] - NVIDIA confirmed its investment in Elon Musk's xAI, contributing $2 billion in a recent funding round, as part of its broader strategy to support AI applications [4] Group 3: Industry Trends and Performance - Hon Hai Precision Industry (Foxconn) reported a 38.01% quarter-over-quarter revenue increase, driven by a surge in AI server shipments [3] - SoftBank is acquiring ABB's robotics unit for $5.375 billion, indicating a strategic shift towards physical AI and automation technologies [7] - Tencent's Hunyuan-Vision-1.5-Thinking model ranked third globally in the latest visual model rankings, highlighting the competitive landscape in AI model development [9] Group 4: Talent Movement in AI Sector - Yao Shunyu, a notable physicist, transitioned from Anthropic to Google DeepMind, citing a desire for more opportunities in the rapidly evolving AI field [10][11]

Core Viewpoint - The recent implementation plan for service-oriented manufacturing by seven departments is not just a policy document but a roadmap for investment opportunities over the next three to four years, emphasizing the integration of manufacturing and services [1][3]. Group 1: Service-Oriented Manufacturing - Service-oriented manufacturing transforms traditional manufacturing by combining product sales with services, such as remote monitoring and maintenance, enhancing efficiency and stability for businesses [3]. - The plan outlines specific goals to be achieved by 2028, including the establishment of 20 standards, the creation of 50 leading brands, and the development of 100 innovation hubs, indicating a clear growth trajectory for the industry [3][4]. Group 2: Investment Opportunities - Companies that can effectively integrate "manufacturing + services" are expected to receive more policy support and capture greater market share, leading to improved performance [3][4]. - The plan highlights the need for adequate computing power infrastructure to support smart manufacturing, suggesting that companies providing such services will have tangible orders and growth potential [4]. - The integration of AI with service-oriented manufacturing is emphasized, with a focus on companies that can deliver practical solutions rather than just concepts [4]. Group 3: Shared Manufacturing - The concept of shared manufacturing, such as shared factories and open testing resources, is gaining attention, allowing smaller companies to access high-end equipment at reduced costs, thus driving growth for companies operating these platforms [4]. Group 4: Long-Term Planning - The policy spans from 2025 to 2028, indicating a long-term commitment to the integration of advanced manufacturing and modern service industries, which are crucial for building a modern industrial system [4][5]. - Companies must demonstrate a genuine commitment to service-oriented manufacturing with real orders and revenue to be considered viable investment opportunities, rather than relying solely on branding [5].

Core Viewpoint - The publishing sub-sector exhibits high dividend attributes and stability within the media sector, with leading companies showing gross margins between 30%-40%, net margins around 10%, and ROE generally above 8% [1] Group 1: Publishing Sector Overview - The publishing sector is characterized by a clear competitive landscape, with at least one publishing group in each province, focusing on both publishing and distribution, including textbooks and supplementary materials as key business areas [1] - The stock price changes in the publishing sub-sector in 2023 are attributed to a market consensus on valuation reassessment, as the content copyrights of publishing companies serve as important sources for data corpus in the context of AI developments [1] - In 2024, the market shows a preference for high-dividend sectors, with leading companies in the publishing sector having relatively high dividend yields compared to the media sector [1] Group 2: Company Analysis - Xinhua Wenhui - Xinhua Wenhui is one of the largest leading companies in the publishing sector, demonstrating outstanding management and operational capabilities [2] - The company's management capabilities are evident in its integrated supply chain services, focusing on both demand and supply-side management, and enhancing content production quality and efficiency [2] - Operational capabilities include developing new growth points through store adjustments and online-offline integration to mitigate external risks, as well as optimizing product structure in response to educational policy changes [2] Group 3: Business Segments - The company has a stable development across various business segments, including 15 publishing media units covering books, periodicals, audio-visual, electronic, and online categories [2] - In reading services, the company operates 181 retail stores in Sichuan Province and has established a multi-scenario online and offline reading service system [2] - The education service network consists of 152 subsidiaries covering Sichuan Province, with clear division of responsibilities between headquarters and subsidiaries [2] Group 4: Investment Outlook - The company is expected to achieve net profits of 1.681 billion, 1.779 billion, and 1.910 billion yuan from 2025 to 2027, with corresponding PE ratios of 11, 10, and 10 times [3] - The company is rated as "recommended" for its strong management and operational capabilities, which are expected to drive steady growth across its business segments [3]

工信部网站10月11日消息，工信部等七部门近日印发《深入推动服务型制造创新发展实施方案（2025—2028年）》，《实施方案》从企业、行业、区域和 生态四个维度，提出体系化推动服务型制造创新发展的任务措施。 其中提到，加强新型信息基础设施建设，深化"5G+工业互联网"融合应用，按需布局算力基础设施，提升工业数据要素供给，推动人工智能技术与服务型 制造融合创新，提升网络和数据安全保障能力。 | 中华人民共和国工业和信息化部 | | | | ■ 阳光小信 无障碍 手机端 邮箱 微信 | 徽博 RSS订阅 | | --- | --- | --- | --- | --- | --- | | Ministry of Industry and Information Technology of the People's Republic of China | | | | 请输入关键字 | Q | | ಿ 首页 组织机构 新闻发布 | | 政务服务 | 政务公开 | 互动交流 | 工信数据 | | 首页 > 工业和信息化部 > 机关司局 > 产业政策与法规司 > 工作动态 | | | | | | | 发文机关： 工业和信息 ...

Macro Dynamics - The Ministry of Industry and Information Technology and six other departments issued a notice to enhance the construction of new information infrastructure, emphasizing the integration of computing power with industry applications and the development of high-quality industry data sets [2] Market Data - In September, the retail sales of passenger cars in China reached 2.239 million units, a year-on-year increase of 6%, and a month-on-month increase of 11%. Cumulatively, retail sales for the year reached 17.004 million units, up 9% year-on-year [5] - The wholesale of passenger cars in September was 2.770 million units, a year-on-year increase of 11%, with a cumulative wholesale of 20.812 million units for the year, up 13% year-on-year [5] Company Dynamics - Didi Autonomous Driving announced a D-round financing of 2 billion yuan, with funds aimed at increasing AI research and development and promoting the application of Level 4 autonomous driving [13] - Wahaha Group appointed Xu Simin as General Manager, while the Chairman position remains vacant following the resignation of Zong Fuli [14] - Hongsheng Beverage Group's legal representative, Zhu Lidan, responded to rumors of being taken away for investigation, urging not to believe in rumors [12]

Core Insights - OpenAI is facing intense internal competition for GPU resources, with a total investment of $7 billion in computing power for 2024, primarily for large model development and inference computing [1][2][12] - Microsoft has launched the world's first GB300 supercomputer, specifically designed for OpenAI, which can significantly reduce the training time for trillion-parameter models from weeks to days [4][6][10] Group 1: Investment and Resource Allocation - OpenAI has spent $5 billion on large model research and $2 billion on inference computing over the past year [1] - The demand for computing power is described as an "endless pit," leading to a critical need for supercomputing expansion and partnerships [2][21] - OpenAI's leadership team has established a clear resource allocation mechanism to manage GPU distribution between research and application teams [15][19] Group 2: Supercomputer Specifications - The GB300 supercomputer features over 4,600 GB300 NVL72 GPUs interconnected via the next-generation InfiniBand network, enabling high data transfer rates and memory capacity [6][8][10] - The system is designed for large-scale AI supercomputing, with a rack-level design that includes 72 GPUs per rack and a total of 37TB of high-speed memory [7][10] - The architecture supports a performance of up to 1,440 PFLOPS using FP4 Tensor Core technology, enhancing the capabilities for AI applications [10] Group 3: Internal Competition and Challenges - OpenAI's internal GPU allocation process is described as a "painful and exhausting" experience, with teams competing fiercely for limited resources [2][12][13] - The allocation of GPUs is critical for productivity, as the number of GPUs directly influences the capabilities of AI applications [19][21] - OpenAI's Chief Product Officer has emphasized the immediate utilization of newly acquired GPUs, highlighting the urgency of resource allocation [21]

Core Insights - OpenAI's total spending on computing resources reached $7 billion last year, primarily for research and experimental runs rather than final training of popular models [1][3][20] - A significant portion of the $5 billion allocated for R&D compute was not used for the final training of models like GPT-4.5, but rather for behind-the-scenes research and various experimental runs [6][18] Spending Breakdown - Of the $7 billion, approximately $5 billion was dedicated to R&D compute, which includes all training and research activities, while around $2 billion was spent on inference compute for user-facing applications [3][5] - The R&D compute spending includes basic research, experimental runs, and unreleased models, with only a small fraction allocated to the final training of models [5][6] Model Training Costs - Researchers estimated the training costs for significant models expected to be released between Q2 2024 and Q1 2025, focusing solely on the final training runs [11][12] - For GPT-4.5, the estimated training run cost ranged from $135 million to $495 million, depending on cluster size and training duration [15] - Other models like GPT-4o and Sora Turbo were estimated using indirect methods based on floating-point operations (FLOP), with costs varying widely [17] Research Focus - The analysis indicates that a large portion of OpenAI's R&D compute in 2024 will likely be allocated to research and experimental training runs rather than directly producing public-facing products [18] - This focus on experimentation over immediate product output explains the anticipated significant losses for OpenAI in 2024, as the company spent $5 billion on R&D while generating only $3.7 billion in revenue [20][21] Power of Compute - The article emphasizes the critical importance of compute power in the AI industry, stating that whoever controls the compute resources will dominate AI [22][28] - OpenAI has engaged in substantial compute transactions, including building its own data centers to mitigate risks associated with reliance on external cloud services [22][30] - The demand for compute resources in AI development is described as having no upper limit, highlighting the competitive landscape [27][28]

Core Viewpoint - Lianyungang City is focusing on leveraging its unique marine resources to create a competitive advantage in emerging industries such as artificial intelligence, computing power, and new energy vehicles, while avoiding homogeneous competition among regions [1] Group 1: Streamlining Assessment - Lianyungang District is reducing the complexity of performance assessments by consolidating multiple evaluation systems into a single comprehensive framework, resulting in a 28% reduction in assessment indicators for rural areas by 2025 [1] - The district is eliminating irrelevant performance indicators and awards that do not align with local realities, such as "Investment Attraction Award" and "Business Environment Optimization Award" [1] Group 2: Shaping Development Focus - The district has introduced "marine content" as a key metric for evaluating development, including a new indicator for the proportion of marine industry investments in newly signed projects [2] - Specific assessments are tailored to different functional areas to avoid homogeneous competition, with a focus on marine power, modern marine fisheries, and coastal tourism [2] - The marine fisheries sector is projected to achieve an added value of 1.866 billion yuan in 2024, with an annual growth rate of 26.3% [2] Group 3: Motivating Performance - Lianyungang District has established a clear incentive structure that rewards high-performing units and penalizes underperformers, promoting accountability among officials [3] - Since 2025, 23 outstanding officials have been promoted, while 3 underperforming officials have been reassigned, effectively enhancing motivation and performance within the district [3]