Core Insights - The strategic value of computing power has transcended mere technology, becoming a key variable in reshaping industry structures, with China's AI-related ICT market investment expected to reach 1.6 trillion yuan by 2025 and 3.5 trillion yuan by 2028 [1] - The rapid evolution of large model technology is driving industries towards digital and intelligent transformation, presenting challenges such as "computing power selection difficulties" and "ambiguous transformation paths" for many enterprises [1][2] - The demand for autonomous innovative computing power is rapidly increasing in the Chinese market, projected to occupy nearly half of the market share by the end of this year [4] Industry Trends - As of June 2023, China's total computing power ranks second globally, with 10.85 million standard racks in use and an intelligent computing power scale of 788 EFLOPS [2] - The traditional IT architecture is struggling to meet the high demands of AI in terms of computing power, data, and real-time processing, leading to the emergence of "data islands" [2] - The shift in computing demand structure from training to inference is driving the rise of autonomous innovative computing power, which is expected to accelerate the industry's move towards intelligence [4] Technological Developments - The new AI transformation era is characterized by the need for computing facilities to possess high-performance multi-core processing capabilities and AI application acceleration technologies [3] - Huawei's computing industry in China is projected to grow by 80% in 2024, with the Kunpeng and Ascend platforms playing crucial roles in various sectors such as government, finance, energy, and transportation [5] Solutions and Frameworks - Huawei has developed an "Industry Intelligent Reference Architecture" to help enterprises find suitable upgrade paths for their digital and intelligent transformation, consisting of five layers: intelligent perception, intelligent connection, intelligent foundation, intelligent platform, and AI models [7][8] - The "Three Layers, Five Stages, Eight Steps" methodology provides a comprehensive roadmap for enterprises, detailing the levels, stages, and specific tasks involved in the transformation process [8] Collaborative Efforts - Huawei aims to leverage partnerships to extend the advantages of Kunpeng and Ascend technologies to a wide range of industries, emphasizing the importance of collaboration in achieving widespread intelligent solutions [9] - The company plans to continue investing in computing technology and strengthen its partner ecosystem to drive innovation and accelerate the digital transformation process across various sectors [9]

作为数字经济的核心引擎，算力的战略价值早已突破单纯的技术范畴，成为重构产业格局的关键变量。 国际数据公司（IDC）预测，中国区2025年仅人工智能相关ICT（信息与通信技术）市场的投资规模就 有望达到1.6万亿元，到2028年更有望增长至3.5万亿元。 "2025年，算力产业迎来千载难逢的发展机遇。以DeepSeek为代表的大模型，为AI产业注入新活力，带 动更庞大的算力尤其是推理需求。"近日，以"先进算力新选择"为主题的2025华为算力场景发布会暨北 京xPN伙伴大会召开，华为商业销售部总裁张海平在会上说。 随着大模型技术的爆发式演进，千行百业正加速迈向数智化转型的深水区。这场数智浪潮下，不少企业 面临"算力选择困难""转型路径模糊"等难题，难以激活先进算力价值。如何破解企业"算力落地难"，推 动先进算力与行业场景的深度融合、最大化释放价值，已成为产业界关注的核心问题。 我国算力总规模位居全球第二。2025中国算力大会发布的数据显示，截至今年6月底，我国在用算力中 心标准机架达1085万架，智能算力规模达788EFLOPS（每秒百亿亿次浮点运算）。 AI变革传统IT架构 "当前数字化和智能化成为经济增长新引 ...

Group 1 - Huawei held the "2025 Huawei Computing Power Scenario Release Conference and Beijing xPN Partner Conference" on August 15, focusing on the theme of "Advanced Computing Power New Choices" [1] - Huawei executives shared advancements in general computing, large model technology, and AI application innovations, highlighting the role of Ascend AI processors and Kunpeng processors across various sectors such as education, government, healthcare, and manufacturing [1] - Huawei's commercial sales president revealed that in 2024, the company's enterprise business and computing industry in China achieved significant growth, with the enterprise business growing by 25% and the computing industry by 80% [1] Group 2 - IDC's vice president analyzed global and Chinese computing power market trends, noting that digitalization and intelligence are becoming new engines for economic growth, with steady growth in ICT spending [2] - Huawei's Ascend computing vice president announced a strategy for comprehensive coverage of edge scenarios from low to high computing power by the second half of 2025, focusing on products like Atlas 300I A2 and 310 modules [2] - Huawei's Kunpeng computing vice president stated that Kunpeng's growth rate is ten times that of x86 computing, positioning it as the preferred choice for independent innovation [2] Group 3 - The "xPN Pioneer Action" was officially launched during the conference, aimed at supporting partners across research, marketing, supply, and service to enhance competitiveness in product and solution development [3] - The conference served as a platform for collaboration in the computing industry, promoting the transformation of advanced computing from technical parameters to practical drivers for industry digital transformation [3] - Huawei emphasized its commitment to deepening collaboration with partners to drive innovation across various industry scenarios using advanced computing power [3]

Core Viewpoint - The article emphasizes the resilience and potential of China's semiconductor industry, highlighting the importance of foundational research and the development of artificial intelligence technologies. Group 1: Chip Development - Ren Zhengfei stated that there is no need to worry about chip development, asserting that many Chinese companies are making significant progress in this area, including Huawei [3] - He acknowledged that Huawei is currently one generation behind the U.S. in single-chip technology but emphasized the use of mathematical methods to compensate for this gap [3] - The potential for growth in mid-to-low-end chips and compound semiconductors in China was highlighted, along with the assertion that software development will not be hindered by external pressures [3] Group 2: Fundamental Research - Ren Zhengfei called for greater societal support for fundamental theoretical research, stressing that without it, China cannot catch up with the U.S. [4] - He noted that Huawei invests approximately 180 billion yuan annually in R&D, with around 60 billion yuan dedicated to fundamental research, which is not subject to performance evaluation [4] - The importance of establishing a strong foundation in technology to ensure long-term growth and stability in the industry was emphasized [4] Group 3: Artificial Intelligence - Ren Zhengfei described artificial intelligence as potentially the last major technological revolution for humanity, with a development timeline spanning decades [6] - He pointed out China's advantages in AI, including a large youth population and robust infrastructure for power and communication networks [6] - The article mentioned the potential for a unified market in China, which could facilitate breakthroughs in technology despite external pressures [6] Group 4: Product Development - The article noted the launch of the Huawei Mate60 series featuring "Chinese chips" and the upcoming release of the HarmonyOS NEXT, which aims for full self-reliance in operating system technology [5] - Huawei's development of various hardware and software products, including Kunpeng processors and Ascend processors, was highlighted as part of its strategy to build a comprehensive computing ecosystem [5]

Core Viewpoint - The dialogue between Ren Zhengfei and People's Daily highlights Huawei's resilience in the face of external pressures, emphasizing a focus on progress rather than obstacles [1]. Group 1: Discussion on Chips - Ren Zhengfei downplays concerns regarding chip technology, stating that many Chinese companies are making significant strides in this area, and Huawei is one of them [3]. - He acknowledges that Huawei is currently one generation behind the U.S. in single-chip technology but believes that through innovative methods, practical results can be achieved [3]. - Ren emphasizes the potential for Chinese companies in mid-to-low-end chips and highlights the importance of compound semiconductors [3]. Group 2: Importance of Basic Research - Ren stresses the necessity of foundational theoretical research for technological breakthroughs, indicating that without it, China will lag behind the U.S. [5]. - He mentions that Huawei invests approximately 180 billion yuan annually in R&D, with around 60 billion yuan dedicated to basic theoretical research [5]. - The company has made significant advancements in "root technology" and operating system innovation, contributing to the foundation of China's information industry [5]. Group 3: Digital Infrastructure and AI - Huawei has launched a series of products, including Kunpeng processors and Ascend processors, to build a computing ecosystem for general and AI applications [6]. - Ren views artificial intelligence as potentially the last major technological revolution, with China having advantages such as a robust power supply and advanced communication networks [8]. - He believes that the future of the country is promising, with increasing openness leading to progress and the potential for a unified market [8].

Investment Rating - The industry rating is "Recommended" [4] Core Insights - The domestic computing power industry is experiencing a pivotal shift from technology following to innovation leadership, driven by the acceleration of AI and digital transformation [13][18] - The performance of domestic chips continues to break through, and the open-source ecosystem is improving, indicating that the domestic computing power industry has entered a new stage of high-quality development [13][18] - The total computing power scale in China is expected to grow by 16.5% year-on-year in 2024, with over 8.8 million standard racks in use [24] Summary by Sections Section 1: Domestic Computing Power - Price Fluctuations and Demand Growth - Domestic computing power card prices are on the rise, driven by increased demand and supply shortages [19][20] - AI technology advancements are leading to a comprehensive price increase in cloud infrastructure, with significant implications for the industry structure [21][22] - Alibaba plans to invest at least 380 billion RMB in AI and cloud computing infrastructure over the next three years, marking a substantial commitment to the sector [27][28] Section 2: Domestic Computing Power - Breaking Through and Leading New Development Trends - Various cities are actively promoting the construction of computing power centers, such as Zhengzhou, Gui'an, and Shenzhen, to meet digitalization demands [36][37] - The upgrade of domestic large models has led to a surge in demand for computing power leasing, with many companies accelerating procurement [41] - Shenzhen has issued nearly 200 million RMB in "training coupons" to support AI and intelligent robotics companies, enhancing the computing power infrastructure [47][48] Section 3: Investment Recommendations - Beneficiary stocks include: 1. AIDC: Runjian Co., Guanghuan New Network, Aofei Data, Runze Technology [17] 2. Computing Power Leasing: Hainan Huatie, Hongjing Technology, Hongxin Electronics [17] 3. Cloud Computing: Youkede, Capital Online, Shunwang Technology, Deepin Technology [17] 4. Computing Power Chips: Haiguang Information, Cambricon, Chipone, Huawei [17] 5. AI Power: Zhongheng Electric, Magmi Tech, Oulu Tong, KOTAI Power [17]