Core Viewpoint - The article highlights the increasing trend of cornerstone investments in Hong Kong IPOs, indicating a strong interest from long-term capital, particularly in new economy enterprises, which is seen as a vote of confidence in the market's future potential [1][14]. Group 1: Cornerstone Investment Trends - Cornerstone investors are crucial in Hong Kong IPOs, committing to purchase large amounts of shares at the issue price and agreeing to lock up their shares for a certain period, typically at least six months [3]. - In 2025, 89 IPO companies are expected to introduce cornerstone investors, with total investments reaching approximately 1,066 billion HKD, marking a historical high [3][5]. - In January 2026, all 13 new IPOs in Hong Kong included cornerstone investors, with total investments amounting to 185.21 billion HKD, a significant increase of 13.3 times compared to the previous year [3][5]. Group 2: Investor Composition - The composition of cornerstone investors includes insurance funds, foreign institutions, and leading private equity firms, with notable participants such as Taikang Life, Temasek, BlackRock, and Morgan Stanley [1][7][8]. - Insurance funds have shown a marked increase in participation, with 15.58 billion HKD in cornerstone subscriptions across 10 IPOs this year, compared to 26.20 billion HKD across 12 IPOs in 2025 [7]. Group 3: Sector Preferences - Long-term capital, including insurance funds and sovereign wealth funds, tends to favor large IPOs and industry leaders, particularly in sectors like artificial intelligence, high-end manufacturing, semiconductors, and biomedicine [10][12]. - New economy leaders are particularly attractive to capital, with significant investments in companies like Zhiyuan, Biran Technology, and others, reflecting a strong interest in sectors aligned with national strategic directions [10][11]. Group 4: Market Dynamics - The active participation of long-term capital in cornerstone investments is viewed as a confidence vote in the ongoing economic transformation and upgrading in China [14]. - The Hong Kong market is becoming a vital window for global capital to allocate Chinese assets, with improved investment quality and market mechanisms attracting significant foreign investment [15].

Group 1 - The Chinese AI industry is entering a new phase of commercial validation and large-scale application, with companies like Zhiyu Huazhang, MiniMax, TianShu ZhiXin, and BiRan Technology recently listing on the Hong Kong Stock Exchange and the Sci-Tech Innovation Board [1] - Domestic chip companies are facing a "bottleneck" dilemma under the ecosystem built by Nvidia, with some listed GPU companies experiencing significant stock price corrections after substantial increases, reflecting market scrutiny of their commercialization paths and long-term growth logic [1] - Since domestic chips cannot quickly catch up with Nvidia in absolute computing power, the focus is shifting to system efficiency and scene applicability, emphasizing "domestic adaptation" to enhance computing power utilization efficiency and accelerate the application of large models across various industry scenarios [1] Group 2 - Industry experts believe that breakthroughs in single-point technology are insufficient for winning the competition, and the collaboration of ecosystems, particularly the "mutual engagement" between models and chips, is crucial for the true autonomy of domestic AI [1] - Looking ahead to 2026, significant breakthroughs in large models are expected in areas such as reinforcement learning, model memory, and context engineering, moving from short context generation to long cognitive chain tasks and from text interaction to native multimodal capabilities, advancing towards the long-term goal of AGI [2] - The willingness of enterprise users to pay for AI-assisted programming tools is expected to increase, benefiting domestic AI large models and facilitating better commercialization as the business value of these tools in software development, data analysis, and business process automation becomes recognized [2] Group 3 - Relevant Hong Kong stocks related to AI large models include MINIMAX-WP (00100), Zhiyu (02513), and Kuaishou-W (01024) [3]

Core Insights - The cornerstone investment market in Hong Kong has seen a significant increase in activity since the beginning of 2026, with strong interest from long-term capital [1] - In 2025, a record 89 IPO companies in Hong Kong attracted cornerstone investors, with a total investment amount of approximately HKD 1,066 billion [1] - In January 2026, all 13 newly listed stocks in Hong Kong attracted cornerstone investors, with total investments reaching HKD 185.21 billion, marking a substantial year-on-year increase of 1,233% [1] Investment Composition - The investor composition includes insurance funds, foreign institutions, leading private equity firms, and industrial capital, all actively participating in cornerstone investments [2] - Insurance funds have shown a notable increase in participation, engaging in cornerstone subscriptions for 10 IPOs in Hong Kong, totaling HKD 15.58 billion [1][2] - In 2025, insurance funds participated in 12 cornerstone subscriptions, amounting to HKD 26.20 billion [1] Market Dynamics - Recent policies have encouraged long-term capital to enter the market, optimizing the investment environment and extending assessment periods, which supports insurance funds in diversifying their investment channels [1] - The increase in IPO activity and the availability of quality investment targets, combined with the pressure on traditional fixed-income asset returns, have driven insurance funds to explore cornerstone investments [1] Focus on New Economy Leaders - New economy leaders have become a focal point for cornerstone investment, with significant amounts raised by companies such as Zhizhu, Biran Technology, and others in January 2026 [2] - Zhizhu, an AI model company, attracted a total of HKD 29.84 billion from multiple institutions, while Biran Technology garnered HKD 28.99 billion, accounting for 63.96% of its global offering [2] - Other companies like MINIMAX-WP, Zhaoyi Innovation, and Haowei Group also saw substantial participation, with each attracting no less than HKD 20 billion from at least 10 institutions [2]

Group 1 - The Chinese AI industry is entering a new phase of commercial validation and large-scale application, with companies like Zhiyu Huazhang, MiniMax, TianShu ZhiXin, and BiRan Technology recently listing on the Hong Kong Stock Exchange and the Sci-Tech Innovation Board [1] - Domestic chip companies are facing a "bottleneck" dilemma under the ecosystem built by NVIDIA, with some listed GPU companies experiencing significant stock price corrections after substantial increases, reflecting market scrutiny of their commercialization paths and long-term growth logic [1] - Since domestic chips cannot quickly catch up with NVIDIA in absolute computing power, the focus is shifting to system efficiency and scene adaptability, emphasizing "domestic adaptation" to enhance computing power utilization efficiency and accelerate the application of large models across various industry scenarios [1] Group 2 - The industry consensus is that breakthroughs in single-point technology are insufficient for winning the competition; ecological collaboration, particularly the "two-way approach" between models and chips, is becoming crucial for the true independence of domestic AI [1] - Looking ahead to 2025, global large model technology capabilities are expected to advance, overcoming productivity scenarios and making significant progress in reasoning, programming, Agentic, and multimodal capabilities, although there are still shortcomings in model generalization stability and hallucination rates [1] - By 2026, further breakthroughs in large models are anticipated in areas such as reinforcement learning, model memory, and context engineering, moving from short context generation to long cognitive chain tasks and from text interaction to native multimodal, progressing towards the long-term goal of AGI [2] Group 3 - The commercial value of AI-assisted programming tools is gradually being recognized, leading to an increased willingness among enterprise users to pay for software development, data analysis, and business process automation scenarios, which is expected to benefit domestic AI large models [2] - Relevant Hong Kong stocks related to AI large models include MINIMAX-WP (00100), Zhiyu (02513), and Kuaishou-W (01024) [3]

然而，回到产业现实中，在英伟达构建的生态高墙下，国产芯片面临的"卡脖子"困境依然存在。部分已 上市GPU公司股价在经历大幅上涨后出现明显回调，一定程度上也反映出市场对其商业化路径和长期成 长逻辑的审视。 既然国产芯片在绝对算力上难以短期追平英伟达，那就从系统效率、场景贴合度上寻求超越。近期芯片 企业和大模型企业的发布中，都在强调"国产适配"，即通过联合优化提升算力利用效率，加速大模型在 各行业场景中的应用落地。 智通财经APP获悉，业内普遍认为，单点技术的突破不足以赢得这场竞争，生态的协同，尤其是模型与 芯片的"双向奔赴"，正成为国产AI能否真正自主的关键。 中金指出，回顾2025年，全球大模型技术能力向前演进，逐步攻克生产力场景，在推理、编程、 Agentic以及多模态等能力方向取得明显进步，但模型通用能力在稳定性、幻觉率等方面仍存在短板。 展望2026年，中金认为大模型在强化学习、模型记忆、上下文工程等方面将取得更多突破，从短context 生成到长思维链任务，从文本交互到原生多模态，并向实现AGI长期目标更进一步。 广发证券此前认为，在AI辅助编程工具给企业带来研发效率提升和业务优化的商业价值逐渐被认识 ...

Core Viewpoint - The stock price of Cambricon has been under pressure due to unverified rumors, despite a recent positive earnings forecast for 2025, indicating a potential shift in market dynamics with increasing competition from new entrants in the AI chip sector [3][5][6]. Stock Performance - On February 3, Cambricon's stock price fell by over 13% during trading, closing down 9.18% at 1128 CNY per share, with a market capitalization of 475.7 billion CNY [5]. - The following day, the stock continued to decline by 5.32%, closing at 1068 CNY per share, resulting in a total market value drop exceeding 70 billion CNY over two days [5]. - Year-to-date, as of February 4, Cambricon's stock has experienced a cumulative decline of 21.21% [6]. Earnings Forecast - Cambricon recently announced an optimistic earnings forecast for 2025, projecting revenue between 6 billion to 7 billion CNY, representing a year-on-year growth of 410.87% to 496.02% [6]. - The company expects to achieve a net profit attributable to shareholders of 1.85 billion to 2.15 billion CNY, marking a turnaround from losses [6]. Market Competition - The entry of new competitors in the AI chip market, such as Moore Threads and MuXi, has diminished Cambricon's previous market scarcity, leading to increased competition and potential pressure on its valuation [8][9]. - The competitive landscape is shifting from a scarcity-driven market to one characterized by increased competition, which may impact Cambricon's growth rates moving forward [8][9]. Industry Dynamics - The AI chip market is experiencing rapid growth, but the sustainability of Cambricon's high growth rates is questioned due to the intensifying competition and the potential for market share dilution [7][8]. - Cambricon's market share in the AI accelerator market is relatively small, with estimates indicating it holds about 1.4% compared to Nvidia's 70% [8][9]. Competitor Performance - Other emerging players in the AI chip sector have shown significant revenue growth, with projections for 2025 indicating substantial increases for companies like Moore Threads and MuXi, which may further challenge Cambricon's market position [9][10].

Core Insights - The Chinese AI industry is entering a new phase of commercial validation and large-scale application, with companies like Zhiyuan Huazhang, MiniMax, and others recently listing on the Hong Kong Stock Exchange and the Sci-Tech Innovation Board [1] - Despite advancements, domestic chip manufacturers face significant challenges due to reliance on NVIDIA's ecosystem, which limits their competitiveness in the market [1][3] - The focus is shifting from achieving absolute computing power to enhancing system efficiency and application relevance, with an emphasis on "domestic adaptation" to improve computational efficiency [1][6] Industry Challenges - The AI application landscape in China has shown remarkable vitality, with models like Qianwen and Zhiyuan GLM performing competitively on benchmarks, yet 99% of AI applications still rely on NVIDIA's infrastructure [3][4] - The entrenched NVIDIA ecosystem, developed over nearly two decades, presents high migration costs for AI companies, complicating the transition to domestic solutions [4] - Domestic chips often struggle with performance and integration issues, leading to a cycle of low adoption and slow ecosystem improvement, which in turn keeps production costs high [4][5] Opportunities for Collaboration - The shift in AI development towards continuous and decentralized inference presents an opportunity for domestic chip manufacturers to differentiate themselves [6] - Collaboration between model and chip developers is essential to address ecological challenges, moving beyond simple hardware deployment to full-stack optimization [6][7] - Initiatives like the "Model-Chip Ecological Innovation Alliance" aim to bridge the technical barriers between chips, models, and platforms, focusing on cost reduction and scalable AI applications [7]

摩尔线程、沐曦股份之外，壁仞科技、天数智芯、燧原科技也争相叩开资本市场的大门。寒武纪等一众 企业披露业绩的同时，百度的昆仑芯、阿里的平头哥接连走到台前。 "小作文"可以呼应投资者高预期下的焦虑，但不能成为股价大跌的唯一由头。 从投资者的视角看，无论是寒武纪，还是紧随其后上市的摩尔线程、沐曦股份，大家都面临着背负高预 期之后，如何兑现、何时兑现的考题。硬件的稳定性、性价比、供应能力都是市场的重要考察因素。 附在股价上的预期，自然比单纯的数字更复杂。营收增长幅度、亏损收窄情况、盈利预期的变化，都可 能变成"小作文"制造焦虑的诱导因素。在这个过程中，稍有风吹草动，让市场产生不安情绪自然在所难 免。 从行业自身的视角看，寒武纪需要正视的不仅是资本市场对于芯片题材祛魅，还有同行竞争加剧、同业 态百花齐放。 寒武纪从最初的"高处不胜寒"，到数家国产芯片企业先后奔赴资本市场，股市上相关概念股的稀缺时代 也将结束。 刚刚交出亮眼的业绩，寒武纪就迎来了猝不及防的大跌。 2月4日，寒武纪股价收跌5.32%，报1068元/股，总市值4504亿元。此前一天，公司股价跌去12.88%，市 值从约5300亿元回落到4563亿元，几乎半天 ...

刚刚交出亮眼的业绩，寒武纪就迎来了猝不及防的大跌。 2月4日，寒武纪股价收跌5.32%，报1068元/股，总市值4504亿元。此前一天，公司股价跌去12.88%，市 值从约5300亿元回落到4563亿元，几乎半天蒸发了超过700亿元。 表面看，导火索是一篇流传不广的"小作文"，公司小范围交流中给出的2026年营收指引仅为200亿元， 远低于市场预期。寒武纪很快辟谣，但难挡股市大跌。 自上市起，高估值的寒武纪一直背负着高预期。这种高预期既有国产替代浪潮的大势之趋，也有AI时 代对于芯片需求的蒸蒸日上。 预期之下，以寒武纪为代表，大量国产芯片企业脱颖而出，成为资本市场耀眼的明星。它们普遍估值 高、上市快、融资规模大，在资本市场起起伏伏，关注度也极高。 "小作文"可以呼应投资者高预期下的焦虑，但不能成为股价大跌的唯一由头。 从投资者的视角看，无论是寒武纪，还是紧随其后上市的摩尔线程、沐曦股份，大家都面临着背负高预 期之后，如何兑现、何时兑现的考题。硬件的稳定性、性价比、供应能力都是市场的重要考察因素。 附在股价上的预期，自然比单纯的数字更复杂。营收增长幅度、亏损收窄情况、盈利预期的变化，都可 能变成"小作文"制造焦 ...

Core Insights - The Chinese AI industry is entering a new phase of commercial validation and large-scale application, with several companies recently listed on the Hong Kong Stock Exchange and the Sci-Tech Innovation Board [1] - Despite advancements, domestic chip manufacturers face significant challenges due to reliance on NVIDIA's ecosystem, which limits their competitiveness and market penetration [3][5] - The shift from centralized training to decentralized inference in AI models presents an opportunity for domestic chips to differentiate themselves through deep collaboration with AI model developers [7][10] Industry Challenges - The dependency on NVIDIA's technology has created a "NVIDIA dependency syndrome," with only a few domestic GPUs able to support a limited number of AI models compared to the vast offerings available globally [3][5] - The lack of a robust ecosystem for domestic chips leads to a cycle of low usage, slow feedback, and high development costs, making it difficult for these chips to gain traction in the market [5][6] - The rapid evolution of AI model architectures necessitates flexible and forward-looking chip designs to avoid obsolescence shortly after production [4][5] Collaborative Efforts - Companies are forming alliances, such as the "Model-Chip Ecological Innovation Alliance," to bridge the technological gaps between chips, models, and platforms, enhancing computational efficiency and application deployment [8] - Major firms like Alibaba and Tencent are pursuing strategies that integrate models, cloud platforms, and chips to achieve systemic advantages in efficiency and cost [9][10] - The focus on dual adaptation between models and chips is seen as a critical path to overcoming existing ecological challenges and enhancing competitiveness in the AI landscape [7][9]