对于国产而言，强势表现标的主要是美股映射模式，结构性定价特点突出。由于ASML、AMAT还处于 新高状态，半导体设备仍将是国产芯片板块的主要情绪点。受制于国内互联网厂商的AI逻辑分歧，叠 加美国数据中心相关标的仍处于盘整期，大数据板块（代表性ETF：大数据ETF华宝（516700））仍处 于横盘波动状态，等待产业链整体的共振机遇。 来源：新浪基金 作者：创业板人工智能ETF华宝（159363）基金经理 曹旭辰 英伟达25日发布的财报显示，该公司2026财年第四季度和全财年营收均创新高，其数据中心业务表现亮 眼。财报显示，该公司2026财年第四季度营收为681亿美元，环比增长20%，同比增长73%；2026财年 营收为2159.4亿美元，同比增长65%；四季度净利润为429.6亿美元，同比增长94%。 最值得关注的一点在于，英伟达对其2027财年第一季度的财务表现做出指引：公司预计2027财年第一季 度毛利率为75%（上下50个基点），好于市场74.6%的预期。英伟达仍保持较高的毛利率，意味着海力 士的HBM并未对英伟达显著涨价，也意味着整体海外数据中心的投入受到大宗商品涨价的影响仍相对 可控。 依托较强的Ag ...

Group 1 - The core viewpoint is that AI capital expenditure continues to exceed expectations, with ongoing catalysts in the application sector, suggesting that the high prosperity of AI may persist. Investors are encouraged to pay attention to communication ETFs (515880), semiconductor equipment ETFs (159516), and technology innovation AI ETFs (589110) [1] - Major global AI model updates are frequent, with significant advancements from companies like Anthropic, Google, and domestic firms such as Zhipu AI and Alibaba, indicating a shift towards differentiated competition and potential for long-term profitability improvement [1] - The leading companies in large models are moving towards a differentiated competition route, with some firms raising prices alongside new model releases, suggesting a transition from "land grab" to a phase of "paying for quality" [1] Group 2 - Capital expenditure from overseas cloud providers for 2025 is projected to reach $410 billion, a 67% year-on-year increase, indicating strong demand for AI infrastructure and benefiting upstream related companies [2] - Guidance for capital expenditure from major tech companies like Google, Amazon, and Meta for 2026 exceeds market expectations, reinforcing the ongoing investment in AI infrastructure [2] - The long-term outlook for domestic alternatives in the AI infrastructure sector is positive, supported by technological competition and policy backing [2]

Group 1 - The AI industry is experiencing rapid advancements, with large models continuously iterating and increasing demand for computing power [3][18][29] - Major AI companies are transitioning from a "low-cost/free user acquisition" phase to a "high-quality paid" phase, significantly boosting computing power demand [4][9][24] - During the holiday period, US AI computing stocks performed exceptionally well, with companies like Lumentum, Corning, and Coherent reaching historical highs [12][28][29] Group 2 - Anthropic launched the Claude Sonnet 4.6 model, which shows significant improvements in programming and reasoning capabilities, becoming the default model for its platforms [4][19] - Google's Gemini 3.1 Pro model outperformed several competitors in benchmark tests, achieving a score of 77.1% on the challenging ARC-AGI-2 test, doubling its predecessor's performance [6][21] - Zhizhu AI released the GLM-5 model, featuring a parameter scale of 744 billion and training data of 28.5 trillion tokens, achieving top scores in open-source model benchmarks [7][22] Group 3 - MiniMax introduced the M2.5 model, designed for agent scenarios, demonstrating significant improvements in programming tasks and cost efficiency [9][24] - Alibaba's Qwen3.5-Plus model outperformed similar open-source models, offering competitive pricing and extensive language support [10][25] - The domestic AI market saw a surge in user engagement during the Spring Festival, with significant increases in active users for various AI applications [12][27]

Market Performance - The market experienced a high opening followed by a pullback, with the ChiNext index rising over 2% at one point. The total trading volume in the Shanghai and Shenzhen markets reached 2.2 trillion, an increase of 219.4 billion compared to the last trading day before the holiday [1] - By the end of the trading day, the Shanghai Composite Index rose by 0.87%, the Shenzhen Component Index by 1.36%, the ChiNext Index by 0.99%, and the CSI A500 Index by 1.20% [1] Consumer Trends - During the recent Spring Festival holiday, consumer activity and travel saw a steady increase, supported by consumption promotion policies. The average daily sales of key retail and catering enterprises increased by 8.6% compared to the same period in 2025 [3][19] - Foot traffic and sales in monitored pedestrian streets increased by 4.5% and 4.8% respectively during the first three days of the holiday compared to last year [3][19] - The focus of economic work for 2026 includes "maintaining domestic demand as the main driver" and "coordinating consumption promotion and investment expansion" [3][19] AI Sector Developments - The AI sector continues to see significant capital expenditure and frequent updates to large models, indicating a sustained high level of activity. Companies like Zhiyuan, ByteDance, Alibaba, and MiniMax have released new AI models, with a focus on application and performance improvements [6][22][23] - The global AI model landscape is evolving, with notable releases such as Anthropic's Claude Sonnet 4.6 and Google's Gemini 3.1 Pro, which enhance capabilities in various tasks [7][23] - Domestic companies are also advancing, with Zhiyuan AI launching its GLM-5 model, which ranks first among open-source models in multiple tests [7][23][24] Infrastructure Investment - The North American AI sector is facing electricity shortages, prompting increased investment in grid infrastructure. The PJM, responsible for the largest regional electricity market in the U.S., plans to enhance its grid to support data centers [10][27] - China's State Grid Corporation is expected to invest 4 trillion yuan during the 14th Five-Year Plan, a 40% increase from the previous plan, focusing on main grid construction and improving cross-province transmission capacity [10][28] - The global energy transition is driving demand for grid construction to accommodate renewable energy, with significant potential in underdeveloped regions [10][28] Robotics Sector Insights - The robotics sector saw high expectations during the Spring Festival, but recent performance has been mixed, with the robotics ETF down by 0.71% [14][30] - Companies showcased advanced capabilities during the Spring Festival, with performances highlighting improvements in motion control and stability [31][32] - Future developments, particularly Tesla's V3 release, are anticipated to impact the robotics market significantly, with varying expectations for its success [32]

Summary of AI Industry Dynamics During the 2026 Spring Festival Domestic Model and Product Releases 1. **Zhiyuan AI** (February 11): Released GLM-5 with a HumanEval code pass rate of 96.2%, ranking first in global open source and fourth overall; focuses on programming and intelligent capabilities [1] 2. **ByteDance** (February 14): Launched Doubao Model 2.0 (Pro/Lite/Mini/Code), achieving top rankings in math/programming benchmarks and reducing inference costs by an order of magnitude; simultaneously launched Seedance 2.0, capable of generating movie-quality multi-shot audio and video in 60 seconds, becoming a key visual creation tool for the Spring Festival Gala [1] 3. **Alibaba** (February 16): Open-sourced Qwen3.5-Plus with a total parameter count of 397 billion and only 17 billion activated, achieving a 60% reduction in memory usage and a 19-fold increase in inference efficiency; API priced at 0.8 yuan per million tokens, which is 1/18 of Gemini 3 Pro [1] Additional Domestic Releases 4. **DeepSeek** (February 17): Upgraded context window from 128K to 1M tokens, capable of processing ultra-long texts equivalent to the "Three-Body Problem" trilogy [2] 5. **MiniMax** (February 18): Released M2.5 model with native agent design, achieving a 37% speed increase for complex tasks [2] 6. **Tencent** (February 19): AI creation reached 1 billion instances over 16 days during the Spring Festival [2] Overseas Model and Product Releases 7. **Google** (February 19): Released Gemini 3.1 Pro, significantly enhancing inference capabilities and setting a new technical benchmark [3] 8. **OpenAI** (February 18): Launched a proprietary model based on Cerebras chips, achieving a training efficiency increase of 5-10 times and a cost reduction of 70% [3] 9. **Anthropic** (February 18): Released Claude Sonnet 4.6, priced at 1/5 of flagship models, with performance approaching Opus, offering excellent cost-performance for enterprise applications [3] 10. **xAI** (February 20): Updated Grok 4.2, with simultaneous upgrades in multimodal and inference capabilities [3] Financing and Capital Dynamics 11. **Moon's Dark Side (Kimi)** (February 20): Completed a $700 million financing round led by Alibaba, Tencent, and Wuyuan, with a valuation exceeding $10 billion; cumulative financing in January and February surpassed $1.2 billion [4] 12. **Zhiyuan AI/MiniMax** (February 20): Market capitalization on the first trading day after the Hong Kong holiday exceeded 300 billion HKD each, totaling over 580 billion HKD [4] 13. **Anthropic** (February 21): Completed a $30 billion Series G round, with a valuation reaching $380 billion, led by GIC and Coatue [4] 14. **OpenAI** (February 21): Nearing completion of the first phase of over $100 billion in financing, with a valuation expected to exceed $850 billion [4] 15. **Runway** (February 21): AI video company completed a $315 million Series E round, with a valuation of $5.3 billion, focusing on world model development [4]

Summary of AI Industry Dynamics During the 2026 Spring Festival Domestic Model and Product Releases 1. **Zhiyuan AI** (Feb 11): Released GLM-5 with a HumanEval code pass rate of 96.2%, ranking first in global open source and fourth overall; focuses on programming and intelligent capabilities [1] 2. **ByteDance** (Feb 14): Launched Doubao Model 2.0 (Pro/Lite/Mini/Code), achieving top rankings in math/programming benchmarks and reducing inference costs by an order of magnitude; simultaneously launched Seedance 2.0, capable of generating movie-quality multi-shot audio and video in 60 seconds, becoming a key player in Spring Festival visual creation [1] 3. **Alibaba** (Feb 16): Open-sourced Qwen3.5-Plus with a total parameter count of 397 billion and only 17 billion activated, achieving a 60% reduction in memory usage and a 19-fold increase in inference efficiency; API priced at 0.8 yuan per million tokens, which is 1/18 of Gemini 3 Pro [2] 4. **DeepSeek** (Feb 17): Upgraded context window from 128K to 1M tokens, capable of processing ultra-long texts equivalent to the "Find Books" trilogy [2] 5. **MiniMax** (Feb 18): Released M2.5 model with native agent design, achieving a 37% speed increase for complex tasks [2] 6. **Tencent** (Feb 19): During the 16 days of the Spring Festival, AI creation reached 1 billion instances [2] Overseas Model and Product Releases 7. **Google** (Feb 19): Released Gemini 3.1 Pro, significantly enhancing inference capabilities and setting a new technical benchmark [3] 8. **OpenAI** (Feb 18): Launched a proprietary model based on Cerebras chips, achieving a training efficiency increase of 5-10 times and a cost reduction of 70% [3] 9. **Anthropic** (Feb 18): Released Claude Sonnet 4.6, with costs at 1/5 of flagship models and performance approaching Opus, offering high cost-performance for enterprise-level applications [3] 10. **xAI** (Feb 20): Updated Grok 4.2, with simultaneous upgrades in multimodal and inference capabilities [3] Financing and Capital Dynamics 11. **Moon's Dark Side (Kimi)** (Feb 20): Completed a $700 million financing round led by Alibaba, Tencent, and Wuyuan, with a valuation exceeding $10 billion; cumulative financing in January and February surpassed $1.2 billion [4] 12. **Zhiyuan AI/MiniMax** (Feb 20): On the first trading day after the Hong Kong stock market holiday, both companies' market capitalizations exceeded 300 billion HKD, totaling over 580 billion HKD [4] 13. **Anthropic** (Feb 21): Completed a $30 billion Series G round, with a valuation reaching $380 billion, led by GIC and Coatue [4] 14. **OpenAI** (Feb 21): Nearing completion of the first phase of financing exceeding $100 billion, with a valuation expected to surpass $850 billion [4]

Core Viewpoint - The artificial intelligence (AI) industry is experiencing a significant surge, driven by the release of new AI models and a shift in focus towards productivity and efficiency in AI applications [1][2]. Group 1: Market Performance - As of February 24, 2026, the CSI Artificial Intelligence Theme Index (930713) rose by 1.29%, with notable gains from stocks such as Runze Technology (up 20.00%), Beijing Junzheng (up 3.52%), and Xiechuang Data (up 2.88%) [1]. - The AI Artificial Intelligence ETF (512930) increased by 1.10%, with the latest price reported at 2.3 yuan [1]. Group 2: Industry Developments - During the Spring Festival, both domestic and international AI developments accelerated, with major companies releasing updated AI models, including ByteDance's Seedance 2.0, Alibaba's Qwen 3.5, and Google's Gemini 3.1 Pro [1]. - The focus of the current AI model updates has shifted from a "parameter competition" to a "productivity competition," emphasizing the ability of models to solve complex real-world tasks [1]. Group 3: Index Composition - The CSI Artificial Intelligence Theme Index comprises 50 listed companies that provide foundational resources, technology, and application support for AI, reflecting the overall performance of AI-related securities [1]. - As of January 30, 2026, the top ten weighted stocks in the index include Zhongji Xuchuang, Xinyi Sheng, and Hanwujing, accounting for a total of 57.27% of the index [2].

Key Insights - The global AI industry is advancing through a collaborative push in computing power, models, and applications, laying a foundation for the scaling of AI video generation technology and the practical deployment of humanoid robots [2][3] - Major AI companies are announcing substantial long-term capital expenditure plans, with OpenAI projecting around $600 billion in computing power spending by 2030, indicating a significant commitment to infrastructure development [3][4] - The recent launch of Seedance 2.0 at the Spring Festival Gala demonstrates the potential of domestic AI models in cultural content creation, marking a significant step towards industrialization in the video generation sector [5][6] Computing Power and Infrastructure - There is a notable acceleration in global AI capital expenditures, with OpenAI initiating a financing round potentially exceeding $100 billion to support its computing power goals [3][4] - In India, Google announced a $15 billion investment in infrastructure, while OpenAI collaborates with Tata to build a 1000 MW data center, showcasing the global expansion of computing resources [3] - Domestic AI companies, such as Moonlight Dark Side, have secured over $1.2 billion in financing, reflecting strong market confidence in Chinese AI enterprises [4] Model Technology - Recent advancements in model technology include the transition to multimodal architectures and cost-effective intelligent models, enhancing performance while restructuring cost frameworks [4][5] - Alibaba's release of the Qwen3.5-Plus model marks a significant upgrade, shifting from pure text training to native multimodal unified modeling, improving performance in long text and multimodal reasoning scenarios [5] - The competitive landscape is shifting from parameter scale competition to architecture efficiency optimization, as demonstrated by Anthropic's Claude Sonnet 4.6 achieving flagship model performance at a lower price point [5] Application Layer - The application of AI in real-world scenarios is advancing, as evidenced by the successful integration of AI video generation technology in the Spring Festival Gala, validating the industrial potential of domestic models [5][6] - Humanoid robots are transitioning from demonstration phases to real-world task execution, with platforms like Volcano Engine and Doubao participating in the Gala, showcasing their capabilities [5] - The AI application landscape is moving towards commercial validation, with content generation and physical execution paths entering a critical phase of market readiness [6] Industry Trends - The AI industry is expected to maintain its upward trajectory, with favorable market conditions supporting a second phase of spring market activity [6] - The successful application of AI during the Spring Festival Gala may act as a catalyst for the current market rally, reinforcing the importance of practical AI applications [6]

Core Insights - The report highlights a significant iteration of large models during the Spring Festival, with advancements in multi-agent collaboration and native multi-modal capabilities driving a leap in performance [1] - Major AI companies have released new foundational models, showcasing features such as parallel agent architecture, complex logical reasoning, and support for ultra-long context [1] - The industry trend is shifting from conversational Q&A to fully automated management of complex engineering tasks [1] Company Developments - Google launched its flagship model Gemini 3.1 Pro, achieving a 77.1% accuracy in the ARC-AGI-2 test and supporting a million-token ultra-long context [2] - Anthropic's Claude Sonnet 4.6 has improved its efficiency in code writing and long text reasoning, slightly outperforming its predecessor Opus 4.6 [3] - xAI introduced the Grok 4.2 model with 500 billion parameters, utilizing a multi-agent cluster mechanism for complex task handling [3] - Alibaba's Qwen 3.5 flagship series integrates linear attention and expert mixture architecture, enhancing decoding throughput by 8.6 times [4] - ByteDance's Doubao 2.0 matrix includes various versions optimized for complex instruction execution, achieving gold medal-level performance in competitions [4] - Zhiyuan AI launched the GLM-5 model with 744 billion parameters, marking a significant advancement in automated intelligent engineering [5] - MiniMax's M2.5 model set industry records in productivity benchmarks, achieving an 80.2% accuracy rate [5] - Kimi's K2.5 model employs joint text-visual pre-training technology, significantly reducing end-to-end reasoning latency [6] Industry Trends - The demand for AI reasoning has led to a price increase in cloud services, with Alibaba Cloud reporting a 34% growth in revenue for Q3 2025, driven by AI-related products [7] - The industry is transitioning from a "price-for-volume" model to a "premium monetization" approach in cloud services [7] - The hardware landscape is shifting from a focus on GPU dominance to a collaborative heterogeneous computing model, with increased demand for CPU and memory due to the rise of AI agents [8] - The need for high concurrency reasoning has highlighted the "memory wall" bottleneck, prompting data centers to adopt high-speed interconnect technologies [8]

Group 1 - The report indicates that non-US assets have continued to exhibit a bullish atmosphere during the holiday period, with risk assets strengthening and stock indices in Europe, East Asia, and South America reaching historical highs [3][4][9] - The performance of overseas risk assets suggests that global liquidity remains abundant, with funds flowing into economies and stock markets expected to show marginal changes, making Chinese stocks attractive to global investors [3][9] - The report maintains the view from the previous report that a new upward cycle is anticipated, encouraging investors to regain confidence and prepare for the first wave of the market's rise in the Year of the Horse [3][53] Group 2 - The report highlights that during the 2026 Spring Festival, major stock indices such as the KOSPI in South Korea and the STOXX 600 in Europe reached new highs, with the KOSPI leading with a 5.48% increase [4][8] - Commodity markets showed a broad upward trend, with NYMEX crude oil surging by 5.94% and COMEX silver rising by 8.63%, indicating strong demand in these sectors [4][8] - The report notes that European corporate earnings have exceeded expectations, with the STOXX 600 index's price-to-earnings ratio at 18.3, significantly lower than the S&P 500's 27.7, suggesting a valuation advantage for European stocks [5][8] Group 3 - The report discusses the AI industry, noting that the demand for high-bandwidth memory (HBM) is driven by the need for massive computing power in generative AI and large model training, with Samsung and SK Hynix dominating the global HBM market [9][48] - The report mentions that the performance of the AI sector is expected to continue to improve, with significant revenue projections for companies involved in AI, such as OpenAI, which anticipates over $280 billion in revenue by 2030 [49][51] - The report highlights the advancements in AI models and the competitive landscape, with major companies like Google and Anthropic releasing new models that enhance capabilities in various sectors [48][49]