Core Viewpoint - The computing power sector, likened to "water, electricity, and coal" in the AI field, has become a new favorite among public funds, with significant increases in holdings of four computing power concept stocks in Q2 2025 [1][2]. Group 1: Fund Holdings and Performance - In Q2 2025, active equity funds significantly increased their holdings in four computing power stocks: Zhongji Xuchuang, Xinyi Sheng, Hudian Co., and Shenghong Technology, with the first two being leaders in optical modules and the latter two being core suppliers of printed circuit boards (PCBs) [1]. - As of June 30, 2025, Shenghong Technology entered the top 20 holdings of public active equity funds, with over 300 funds holding a combined market value of approximately 13.2 billion yuan, accounting for about 11.5% of its circulating shares [3]. - The stock price of Shenghong Technology has surged, with a cumulative increase of over 300% since the beginning of the year, and over 1000% since its low in early February 2024 [1][2]. Group 2: Market Trends and Drivers - The high demand for AI chips, particularly GPUs, has directly catalyzed the stock price of Shenghong Technology, as the GPU market's growth is expected to drive demand for optical modules and PCBs [2]. - The computing power sector is experiencing a high boom, with significant increases in the number of funds holding stocks in this sector, particularly for Xinyi Sheng and Zhongji Xuchuang, which have also seen substantial increases in their stock prices [8][9]. Group 3: Future Outlook - Fund managers express optimism about the computing power sector, noting that while cloud computing has seen significant gains, adjustments and volatility are expected in the future [11]. - The ongoing development of artificial intelligence is anticipated to sustain high demand for the computing power supply chain, with expectations of continued growth in the sector driven by advancements in AI applications and increased capital expenditure [12][13]. - There is a strong belief in the potential of Chinese companies to innovate in foundational models, core algorithms, and advanced chips, which could lead to significant advancements in the global AI landscape [14].

Core Insights - The 2025 Green Computing (AI) Conference was held in Hohhot, Inner Mongolia, focusing on building a green computing ecosystem and creating a new future for artificial intelligence [1] - The China Academy of Information and Communications Technology released a report indicating that China's computing industry is transitioning to a high-quality development phase with a focus on sustainability [1] Group 1: Green Computing Development - Inner Mongolia and the Hohhot New Area have emerged as leaders in green computing development, achieving a data center PUE value as low as 1.15 and a green electricity usage rate exceeding 80% [1] - The report covers the entire process of computing production, supply, operation, and application, summarizing the latest trends and advanced practices in promoting green and low-carbon computing [1] Group 2: Computing Equipment and Infrastructure - China is advancing chip architecture innovation and deploying high-density liquid-cooled servers, with a total computing power expected to reach 280 EFLOPS by the end of 2024, reflecting an average growth rate of nearly 30% over the past five years [2] - The number of standard racks in operational data centers across the country is expected to exceed 9 million by the end of 2024, with hub node areas accounting for over 60% of new computing power [2] Group 3: Energy and Computing Synergy - In 2024, the electricity consumption of data centers is projected to reach 166 billion kWh, representing about 2% of total social electricity consumption, with total carbon emissions from data centers at 0.859 million tons [2] - Strategies include positioning computing power near energy-rich areas and promoting green electricity supply, energy storage, and microgrid models [2] Group 4: Utilization and Policy Initiatives - The implementation of computing vouchers and scheduling platforms has improved computing power utilization rates, with 22 provincial-level administrative regions having launched local computing voucher policies, distributing a total of 1.53 billion yuan annually [3] - Hohhot has established a dual-driven industrial model of "green electricity + computing equipment," with a total computing power of 101,000 P, expected to exceed 120,000 P by the end of 2025 [3]

深圳商报·读创客户端首席记者 王海荣 通讯员 鹏实 7月21日，记者从鹏城实验室获悉，在国家数据局的大力支持下，该实验室积极参与并承担的全国一体化算力监测任务的技术攻关和研发落地取得新进展。 深入实施"东数西算"工程 算力作为数字经济时代的核心生产力，是人工智能技术与产业发展的重要引擎。随着新一轮科技革命加速演进，大规模算力及基础设施将成为大国博弈的战 略制高点，并逐步在提升国家竞争力中发挥关键作用。2023年，国家发展改革委、国家数据局等部门联合印发了《深入实施"东数西算"工程 加快构建全国 一体化算力网的实施意见》，明确提出构建"全国一体化算力网"的目标，通过网络连接多源异构、海量泛在的算力，优化全国算力资源配置。 P C LIMIN 1404 服务人民 标题图 6 图为鹏城实验室。 深圳商报·读创客户端首席记者 王海荣 摄 基于此，国家数据局于2024年10月组织开展了数据基础设施建设（先行先试）任务，探索"全国一体化算力网"实践路径。鹏城实验室作为网络通信领域的国 家战略科技力量，响应国家需求，在国家数据局的大力支持下，积极参与并承担先行先试任务中全国一体化算力监测任务的技术攻关和研发落地。 建设全国 ...

Core Viewpoint - The report highlights the performance and outlook of the Manulife Transformation Opportunity Stock A fund, indicating strong growth in net asset value and a favorable investment environment in the technology sector, particularly among internet giants [2][12]. Fund Performance - The fund reported a profit of 275 million yuan in Q2 2025, with a weighted average profit per fund share of 0.2308 yuan [2]. - The net asset value growth rate for the fund was 10.06% during the reporting period, with a total fund size of 3.028 billion yuan as of the end of Q2 [2][14]. - As of July 18, the fund's unit net value was 3.055 yuan, with a one-year compounded net value growth rate of 31%, ranking it highest among its peers [2]. Comparative Analysis - The fund's performance over the last three months showed a compounded net value growth rate of 40.26%, ranking 2 out of 62 comparable funds [2]. - Over the last six months, the compounded net value growth rate was 16.51%, ranking 12 out of 62 [2]. - The fund's three-year Sharpe ratio was 0.2883, placing it 25 out of 50 in its category [7]. Risk Metrics - The fund experienced a maximum drawdown of 56.63% over the last three years, ranking 5 out of 50 comparable funds [9]. - The maximum drawdown in a single quarter occurred in Q1 2021, reaching 29.7% [9]. Investment Strategy - The fund maintains a high average stock position of 90.72% over the last three years, compared to the industry average of 88.41% [12]. - The fund's top ten holdings have consistently represented over 60% of its portfolio for nearly two years, with significant investments in companies like Xinyi Technology and Huadian Electric [16].

Core Viewpoint - The conference held on July 19 in Xining, Qinghai Province, announced significant growth in green computing power, highlighting Qinghai's role as China's first green computing electricity pilot province [1] Group 1: Growth in Computing Power - Qinghai Province's green computing power increased from 370P (PFLOPS) at the beginning of 2024 to 15,000P by June 2025, representing a nearly 40-fold growth [1] - The province has constructed approximately 190,000 standard racks, with a utilization rate exceeding 90% and a server online rate close to 100% [1] Group 2: Future Projections - The province aims to double both the rack scale and computing power level by 2025 [1] - The current infrastructure includes a comprehensive computing power supply system consisting of "general computing + intelligent computing + supercomputing" [1]

截至2025年7月18日 10:43，中证人工智能产业指数(931071)强势上涨1.53%，成分股深信服(300454)上涨 6.95%，中际旭创(300308)上涨5.67%，金山办公(688111)上涨4.99%，豪威集团(603501)，瑞芯微 (603893)等个股跟涨。人工智能ETF(515980)上涨1.62%， 冲击6连涨。 截至7月17日，人工智能ETF近1年净值上涨44.77%，指数股票型基金排名434/2917，居于前14.88%。从 收益能力看，截至2025年7月17日，人工智能ETF自成立以来，最高单月回报为30.38%，最长连涨月数 为3个月，最长连涨涨幅为43.97%，上涨月份平均收益率为6.80%。 流动性方面，人工智能ETF盘中换手2.71%，成交8910.31万元。拉长时间看，截至7月17日，人工智能 ETF近1周日均成交2.20亿元。 规模方面，人工智能ETF最新规模达32.43亿元。 人工智能ETF(515980)，场外联接(华富人工智能ETF联接A：008020；华富人工智能ETF联接C： 008021)。 人工智能ETF紧密跟踪中证人工智能产业指数，中证人工智能产业 ...

Group 1 - The manufacturing cycle has been reduced from 60 months to 24 months through the integration of "computing + network" technologies, leading to significant upgrades in the entire process from research and development to production [1] - The importance of computing infrastructure has been emphasized as a key facility for industry digital transformation, with the government implementing policies to promote high-quality development in the computing industry [3] - The Ministry of Industry and Information Technology and other departments have released action plans to establish a monitoring mechanism for computing infrastructure and to create benchmark cities for computing networks by 2025 [3] Group 2 - Zhejiang province is the only region in China with a "one center, two direct connection points" model, enhancing its computing capabilities and completing a low-latency computing circle [3] - Hangzhou has made significant progress in expanding its 1ms latency circle, covering over 20 million people and meeting the demands for computing applications through advanced network technologies [4] - The China (Hangzhou) Computing Town has gathered key enterprises focusing on integrated circuit design, high-end software, and intelligent hardware, aiming to build a comprehensive industrial ecosystem [6] Group 3 - The Zhejiang Water Science Culture Group utilizes low-latency computing networks to enable real-time generation of short videos across 500 application scenarios, supporting the development of the cultural tourism industry [7] - Major telecommunications companies in Zhejiang have shared insights on the operational status and unique applications of the city's computing network, highlighting its role in facilitating intelligent transformation in industries like smart manufacturing [4][7]

Summary of Conference Call Records Industry or Company Involved - Focus on the AI computing power sector and its implications for investment opportunities in North America and globally [1][2][3][4][28] Core Points and Arguments 1. **AI Computing Power Demand**: The demand for AI computing power remains strong, with significant capital expenditures from major North American tech companies like Amazon, Microsoft, Google, and Meta, totaling $77.3 billion in Q1, a 62% year-over-year increase [2][3]. 2. **Capital Expenditure Projections**: MECA has revised its annual capital expenditure forecast from $60-65 billion to $64-72 billion, indicating strong optimism in the sector [3][4]. 3. **Token Consumption Growth**: The consumption of tokens, which is closely tied to AI computing power, is expected to grow exponentially, driven by both training and inference processes in AI models [5][6][10][11]. 4. **Model Complexity and Token Demand**: The complexity of AI models, particularly in multi-agent systems, leads to a significant increase in token consumption, with predictions of a 100-fold increase in token processing for single user queries over the next two years [9][10][15]. 5. **Market Dynamics**: The rapid growth in token consumption raises concerns about the sustainability of business models and the potential for market consolidation, where only a few models may dominate the market [12][13][14]. 6. **Investment Sentiment**: Despite the strong demand for AI computing power, there is uncertainty regarding future investments and the potential for a slowdown in capital expenditures if commercial viability is not established [28][42]. 7. **AI Agent Development**: The development of AI agents is seen as a critical area for future growth, with a focus on enhancing their capabilities through memory, planning skills, and tool usage [30][31][33]. 8. **Historical Context**: The discussion includes historical cycles of investment in AI and computing power, suggesting that current trends may lead to significant future growth, albeit with caution due to market volatility [22][24][27][42]. Other Important but Possibly Overlooked Content 1. **Technological Advancements**: The advancements in AI models, particularly in multi-modal capabilities, are expected to enhance the efficiency and effectiveness of AI applications [32][33]. 2. **Telecom Sector Performance**: The telecom sector is experiencing slow growth, with a focus on improving broadband penetration and the potential for increased revenue from smart home services [35][36][39]. 3. **Cash Flow Concerns**: There are concerns regarding the decline in free cash flow among telecom operators, which may impact their ability to sustain capital expenditures in the future [38][39][40]. 4. **Investment Strategy**: The recommendation is to selectively invest in high-potential stocks within the AI sector while maintaining a cautious outlook on overall market conditions [29][42]. This summary encapsulates the key insights from the conference call, highlighting the ongoing developments in the AI computing power sector and the associated investment landscape.

Group 1 - The 2025 Green Computing (Artificial Intelligence) Conference was successfully held in Hohhot, highlighting the city's unique advantages in "green electricity + computing power" as a core engine for high-quality development in the computing power industry [1] - The conference focused on themes such as the integration of green energy and computing power, AI technology innovation, and the systematic promotion of the "East Data West Computing" project [1][4] - Various stakeholders, including government representatives, industry experts, and enterprise representatives, gathered to discuss the upgrading of the computing power industry and explore new pathways for high-quality development [1][4] Group 2 - The construction of computing network infrastructure is crucial for the market-oriented allocation of data elements, with a focus on creating a faster and greener computing network foundation [2] - Companies like H3C Technologies and China Telecom discussed advancements in "green computing networks," AI technology, and quantum technology, emphasizing innovation and collaboration in strategic emerging industries [2] - Beijing Zhongke Xianlu Intelligent Computing Technology Co., Ltd. shared challenges and innovations in data center lifecycle management, establishing a liquid cooling testing laboratory for next-generation data centers [2] Group 3 - The exponential growth of computing power presents significant challenges in data center security and energy supply, necessitating enhanced risk management capabilities [3] - A cooperation agreement was signed among Inner Mongolia, Heilongjiang, Jilin, and Liaoning provinces to promote deep collaboration in computing networks and data sharing [3] - The launch of the first "East Data West Computing" green computing industry risk solution aims to provide comprehensive risk support for the construction of computing power hubs in the region [3] Group 4 - AI computing power is reshaping the digital economy landscape, with companies like Huawei and China Mobile leading discussions on the application of AI computing power in various industries [4] - Huawei introduced the "Star River AI High-Performance Data Center Network" solution to enhance computing power release and support intelligent transformation in sectors like autonomous driving and coal mining [4] - The series of meetings established a collaborative innovation system that integrates policy, technology, and industry, marking a new phase in the development of green computing power in China [4] Group 5 - Hohhot is positioned as a digital hub on the new western land-sea corridor, leveraging the "China Northern Ecological Computing Valley" to drive deep integration between the digital economy and the real economy [5] - The city aims to play a key role in the global digital Silk Road through the construction of a computing power network along the Belt and Road Initiative [6] - The rise of a trillion-level computing power industry cluster in Northern Xinjiang is contributing to the integration of green development and technological innovation in the broader context of building a digital China [6]

Group 1 - The 2025 Green Computing Power Conference was successfully held in Hohhot, Inner Mongolia, showcasing new achievements in China's digital economy and outlining a new blueprint for the development of green computing power [1] - The construction of a national integrated computing power network is a significant marker of modernization, with the Ministry of Industry and Information Technology promoting a system for efficient and convenient computing power utilization [2][3] - Hohhot has established a 400G high-speed optical transmission network connecting with other computing power hubs, ensuring a layered latency guarantee system [2] Group 2 - The integration of green concepts into computing power development is essential for achieving high-quality growth, with policies promoting the synergy between computing power and renewable energy [4][5] - Hohhot is leveraging its advantages to transform into a green computing power hub, addressing high energy consumption issues in data centers while providing scalable applications for the renewable energy industry [5] - The demand for computing power is significantly driven by the proliferation of artificial intelligence technologies, with over 1.3 million computing power application projects reported by the Ministry of Industry and Information Technology [6][7] Group 3 - The implementation of local "computing power vouchers" has been adopted by over 20 provinces, with a total annual issuance amounting to 1.53 billion yuan [7] - Hohhot is focusing on building a comprehensive industrial ecosystem that integrates green electricity and computing power equipment, while also enhancing its capabilities in various key industries [7] - The future of the computing power industry lies in deep integration with data and algorithms, accelerating penetration into various sectors such as transportation, finance, education, healthcare, technology, and energy [7]