Investment Rating - The industry investment rating is "Overweight" [6] Core Views - The report highlights that the server CPU market is experiencing supply shortages due to increased demand driven by AI advancements and cloud computing needs. The delivery cycle for some CPU models may extend up to 6 months [1] - The rapid development of Agentic AI is expected to significantly increase the demand for CPUs, as these systems require more computational power for tasks such as database retrieval and external tool calls [3] - The report emphasizes the importance of CPU in AI systems, predicting that the number of tasks executed by agents will grow from 44 billion in 2025 to 415 trillion by 2030, representing a compound annual growth rate of 524% [3] Summary by Sections Demand 1: Replacement Cycle and AI Demand - The general server market is currently in a 3-5 year replacement cycle, with significant growth in server procurement driven by the rise of remote work and online services during 2020-2021. The global shipment of general servers is expected to grow by 13% in 2024 and 8% in 2025, supporting a robust CPU market [2] Demand 2: Agentic AI Driving CPU Growth - Agentic AI, which integrates large language models and decision orchestration, is expected to create a new wave of demand for CPUs. The report cites a study indicating that CPU performance may become a bottleneck in Agentic AI systems, with tool processing delays accounting for up to 90.6% of total latency [3] Supply: Insufficient Advanced Process Capacity - Intel is facing supply shortages due to inadequate capacity planning for advanced processes, with shortages expected to persist until 2026. The company is prioritizing server CPU supply, but adjustments will not be completed until the end of Q1 2026 [4] Key Recommendations - The report recommends stocks such as Haiguang Information (688041 CH) with a target price of 291.00 and Lanke Technology (688008 CH) with a target price of 178.80, both rated as "Buy" [8][19]

Group 1: Market Performance - On February 12, chip stocks strengthened, with Jingchen Co. leading a 16% increase, followed by Yuanjie Technology, Shengke Communication-U, and Shijia Photonics rising over 5% [1][7] - The "Chip" ETF Huabao (589190) opened high and rose by 1.37% during the session [1][7] Group 2: Company Performance - Jingchen Co. announced a revenue forecast of 6.77 billion for 2025, representing a year-on-year growth of 14.6%, and a net profit of 870 million, up 6% year-on-year [2][10] - The company expects a revenue growth of 10%-20% in Q1 2026 and 25%-45% for the entire year, driven by strong partnerships with major internet companies like Google, Amazon, and META [2][10] Group 3: Industry Insights - The semiconductor industry is currently in an upward cycle, with AI as the core driver; North America's four major cloud providers are expected to increase capital expenditures by 67% year-on-year in Q4 2025 [3][10] - Global semiconductor sales increased by 37.1% year-on-year in January 2026, marking 26 consecutive months of positive growth, with DRAM and NAND Flash spot prices rising approximately 39% and 35% respectively [3][10] Group 4: ETF Performance - The Huabao Chip ETF has an annualized return of 17.93% since its inception, outperforming similar indices [12][13] - The index has shown a maximum drawdown that is smaller compared to other semiconductor indices, indicating a better risk-reward ratio [12][13]

Core Insights - Zhipu AI has launched and open-sourced GLM-5, with multiple domestic chip manufacturers completing Day-0 adaptation for the model, ensuring compatibility and stable operation from the first day of release [1][2] Group 1: GLM-5 Launch and Adaptation - Zhipu AI's GLM-5 has been adapted by major domestic chip platforms including Huawei Ascend, Moore Threads, Cambricon, Kunlun, Muxi, Suiruan, and Haiguang, achieving high throughput and low latency on domestic computing clusters [2] - Haiguang's DCU team collaborated closely with Zhipu AI to optimize the underlying operators and hardware acceleration, enabling GLM-5 to run stably on Haiguang DCU with high throughput and low latency [1] - Moore Threads completed full-process adaptation and verification on its flagship AI training and inference GPU MTT S5000, leveraging its MUSA architecture to enhance model inference performance while reducing memory usage [1] Group 2: Model Performance and Features - GLM-5 has expanded its parameter scale from 355 billion (activated 32 billion) to 744 billion (activated 40 billion) and increased pre-training data from 23 terabytes to 28.5 terabytes, significantly enhancing its general intelligence capabilities [2] - The model incorporates a new "Slime" asynchronous reinforcement learning framework, supporting larger model scales and more complex tasks, and utilizes DeepSeek Sparse Attention to maintain long text performance while reducing deployment costs and improving token efficiency [2] - A month prior, Zhipu AI released the GLM-Image model, which employs a hybrid architecture of autoregressive and diffusion decoders, marking a significant exploration in cognitive generative technology [2]

Core Insights - Zhipu AI has launched and open-sourced GLM-5, with multiple domestic chips completing Day0 adaptation for optimized performance [1][2] - GLM-5 has achieved state-of-the-art (SOTA) performance in coding and agent capabilities, with significant improvements in parameter scale and pre-training data [2] Group 1: GLM-5 Launch and Adaptation - Zhipu AI's GLM-5 has been adapted for high throughput and low latency on domestic chip platforms including Huawei Ascend, Moore Threads, and others [2] - The collaboration with Haiguang Information has optimized the underlying operators and hardware acceleration for stable operation on Haiguang DCU [1] - Moore Threads has completed full-process adaptation and verification on its MTT S5000 GPU, achieving high-performance inference while reducing memory usage [1] Group 2: Technical Enhancements and Performance Metrics - GLM-5's parameter scale has expanded from 355 billion (with 32 billion activated) to 744 billion (with 40 billion activated), and pre-training data has increased from 23 terabytes to 28.5 terabytes [2] - The introduction of the "Slime" asynchronous reinforcement learning framework supports larger model scales and complex tasks, enhancing learning from long-term interactions [2] - The integration of DeepSeek Sparse Attention mechanism significantly reduces deployment costs while maintaining long text performance [2] Group 3: Support from Huawei - Huawei's Ascend NPU and MindSpore AI framework provide comprehensive support from data to training, contributing to the development of SOTA models based on self-innovated computing power [3]

Group 1 - The semiconductor ETF (159813) opened with a gain of 0.57%, priced at 1.225 yuan [1] - Key holdings in the semiconductor ETF include: - SMIC up 0.47% - Haiguang Information up 1.96% - Cambrian up 0.10% - North Huachuang up 0.53% - Zhaoyi Innovation up 2.08% - Zhongwei Company up 0.62% - Lanke Technology up 1.05% - OmniVision up 0.03% - Tuojing Technology up 0.89% - Changdian Technology up 0.67% [1] - The performance benchmark for the semiconductor ETF is the National Securities Semiconductor Chip Index return, managed by Penghua Fund Management Co., Ltd. [1] Group 2 - Since its establishment on April 17, 2020, the semiconductor ETF has achieved a return of 82.59% [1] - The ETF has experienced a slight decline of 0.02% over the past month [1]

Investment Rating - The report maintains a "Positive" investment rating for the industry [7] Core Insights - The computer sector experienced a significant decline of 3.53% last week, ranking 29th among major industries, with a trading volume accounting for 6.50% of the total market [2][4] - The recent surge in demand for AI applications has highlighted a shortage of computing power, prompting major AI companies like Alibaba to increase their procurement of computing resources [6][45] - Key recommendations include focusing on domestic computing chip manufacturers, particularly industry leaders like Cambricon and Haiguang Information, as well as companies involved in computing power leasing and IDC [6][45] Summary by Sections Market Performance - The computer sector's performance was notably poor, with a 3.53% drop, while the overall market index fell by 1.27% [4][14] - Active sectors included cybersecurity and financial IT, with notable stock performances from companies like High凌信息 and 御银股份 [16] Key Developments - ByteDance launched its latest video model, Seedance 2.0, which supports video generation from text and images, marking a significant advancement in AI video technology [19][21] - Breakthroughs in quantum networking were achieved by researchers at the University of Science and Technology of China, enabling long-distance quantum communication [26][35] - The initiation of national standards for intelligent connected infrastructure is expected to accelerate the development of smart transportation systems [36][43] Investment Opportunities - The report emphasizes the investment potential in computing infrastructure due to the recent surge in demand highlighted by the 千问 app's promotional activities, which led to system outages due to overwhelming user traffic [6][45] - The ongoing competition among major tech firms during the Spring Festival is expected to further drive the demand for computing resources [49][53]

Core Insights - The article provides an overview of the latest net asset values of various funds, highlighting the top-performing and bottom-performing funds in terms of net value growth [1][4]. Fund Performance Summary Top 10 Funds by Net Value Growth - The top-performing funds include: 1. Guotai Jinlong Industry Mixed Fund: Latest net value 0.4589, growth of 4.53% 2. Huaxia Core Growth Mixed A: Latest net value 0.8218, growth of 4.21% 3. Huaxia Core Growth Mixed C: Latest net value 0.7980, growth of 4.20% 4. Hui'an Quantitative Pioneer Mixed C: Latest net value 1.7332, growth of 4.16% 5. Hui'an Quantitative Pioneer Mixed A: Latest net value 1.7887, growth of 4.16% 6. GF Quality Preferred Mixed A: Latest net value 1.9460, growth of 4.14% 7. GF Quality Preferred Mixed C: Latest net value 1.9306, growth of 4.14% 8. Zhongyou Cycle Selected Mixed A: Latest net value 1.0857, growth of 4.00% 9. Zhongyou Cycle Selected Mixed C: Latest net value 1.0849, growth of 4.00% 10. Taixin Development Theme Mixed: Latest net value 1.9970, growth of 3.96% [2][4]. Bottom 10 Funds by Net Value Growth - The underperforming funds include: 1. E Fund Vision Growth Mixed A: Latest net value 2.2459, decline of 3.29% 2. E Fund Vision Growth Mixed C: Latest net value 2.2020, decline of 3.29% 3. E Fund Pioneer Growth Mixed C: Latest net value 2.4436, decline of 3.24% 4. E Fund Pioneer Growth Mixed A: Latest net value 2.4871, decline of 3.24% 5. Qianhai Kaiyuan Shanghai-Hong Kong Deep Strong Domestic Industry Mixed: Latest net value 1.7599, decline of 2.63% 6. Fortune Emerging Industry Stock C: Latest net value 3.7914, decline of 2.58% 7. Fortune Emerging Industry Stock A: Latest net value 3.8720, decline of 2.58% 8. GF Emerging Growth Mixed A: Latest net value 1.7701, decline of 2.58% 9. GF Emerging Growth Mixed C: Latest net value 1.7378, decline of 2.57% 10. Zhongjia Technology Innovation Mixed C: Latest net value 2.2426, decline of 2.57% [3][4]. Market Analysis - The Shanghai Composite Index showed slight fluctuations, closing with a minor increase, while the ChiNext Index opened lower and subsequently declined. The total trading volume reached 2 trillion yuan, with a ratio of advancing to declining stocks at 2050:3241 [6]. - Leading sectors included building materials, chemical fibers, and non-ferrous metals, each with gains exceeding 2%. Conversely, the media and entertainment sector experienced a decline of over 3% [6]. Fund Holdings Overview Guotai Jinlong Industry Mixed Fund - The fund's top holdings include: - Haiguang Information: 6.56% of the portfolio, with a daily increase of 5.47% - Yahua Group: 6.25% of the portfolio, with a daily increase of 5.47% - Xiyang Co.: 6.01% of the portfolio, with a daily increase of 2.26% - The fund's holdings are concentrated in resource sectors, with a total holding concentration of 52.83% [7][8]. E Fund Vision Growth Mixed A - The fund's top holdings include: - Zhongji Xuchuang: 9.66% of the portfolio, with a daily decline of 4.28% - Xinyi Sheng: 9.37% of the portfolio, with a daily decline of 5.46% - Yingweike: 8.63% of the portfolio, with a daily decline of 4.13% - The fund's holdings are concentrated in the artificial intelligence sector, with a total holding concentration of 60.86% [8].

Investment Rating - The report maintains a "Buy" rating for the company, with a target price of 208.73 CNY per share, compared to the current price of 180.30 CNY [3]. Core Insights - The company is expected to achieve a total revenue of 420 million to 450 million CNY in 2025, representing a year-on-year growth of 13.06% to 21.13%. The net profit attributable to shareholders is projected to be between -250 million to -220 million CNY [7]. - In Q4 2025, the company anticipates revenue of 190 million to 220 million CNY, with a year-on-year growth rate of 20.4% to 38.9%, significantly higher than the 7.4% growth rate in the first three quarters of 2025 [9]. - The demand for the company's AI foundational software products, such as TDH and Sophon LLMOps, is robust, reflecting a strong market need for digital transformation solutions [9]. - The company is actively investing in the research and development of new AI database technologies, which are expected to enhance its competitiveness in the AI foundational software sector [9]. - The ArgoDB joint solution, based on Haiguang Information's CPU, has been successfully implemented in key industries such as finance, energy, and government, indicating strong market acceptance [9]. Financial Forecast and Investment Recommendations - Revenue projections for 2025 to 2027 are estimated at 440 million, 500 million, and 580 million CNY, respectively, with growth rates of 17.3%, 15.8%, and 14.8% [10]. - The net profit attributable to shareholders is expected to improve from -240 million CNY in 2025 to -70 million CNY in 2027, with corresponding EPS estimates of -1.99 CNY, -0.87 CNY, and -0.58 CNY [10]. - The company is compared to peers such as MongoDB, Snowflake, and Elastic, with a premium valuation due to the scarcity of A-share big data companies and the promising future of AI software tools [15][14].

Core Insights - The article highlights the official launch and open-sourcing of GLM-5 by Zhipu AI on February 11 [1] - Haiguang DCU has completed Day0 adaptation and joint fine-tuning for GLM-5, leveraging its self-developed AI software stack and open ecosystem capabilities [1] - The initiative aims to provide ready-to-use deployment solutions for global developers and enterprise users, facilitating the rapid implementation of GLM-5's core capabilities [1]

Core Viewpoint - Jianghuai Automobile (600418.SH) has raised approximately 3.5 billion yuan through a private placement, with notable participation from prominent investor Ge Weidong, who invested 1 billion yuan [1][10]. Group 1: Fundraising Details - The final pricing for the share issuance was set at 49.88 yuan per share, with a total of 70,168,404 shares issued, raising around 3.5 billion yuan [1][2]. - The funds raised will primarily be used for high-end intelligent pure electric passenger vehicle projects [3][8]. - All shares subscribed by the investors will be locked for six months post-issuance [3]. Group 2: Investor Insights - Ge Weidong's investment of approximately 1 billion yuan for 2,004,811 shares highlights confidence in Jianghuai's collaboration with Huawei in the high-end intelligent electric vehicle sector [10]. - The market is optimistic about the "smart car + AI" model, which is seen as entering a product explosion phase [3][10]. - Ge Weidong's investment strategy focuses on long-term trends, particularly in AI and automotive sectors, indicating a strong belief in Jianghuai's potential [10][14]. Group 3: Company Performance and Future Outlook - Jianghuai Automobile has delivered a total of 12 million vehicles globally and anticipates a sales volume of 384,100 vehicles in 2025, reflecting a year-on-year decline of 4.72% [3]. - The company expects a net loss of approximately 1.68 billion yuan for 2025, attributed to increased competition in the international market and losses from joint ventures [9]. - The company plans to launch three new models under the "Zun Jie" brand in 2026, aiming to strengthen its position in the ultra-luxury market [6][8].