| 昌亮 | 214.000 | 最低 | 155.000 | | --- | --- | --- | --- | | 開市 | 155.000 | 前收 | 144.600 | | 成交量 | 6700 | 成交額 | 140.04萬 | | 平均價 | 209.008 | 振幅 | 40.80% | | 總市值 → | 534.07億 | 總股本 | 2.54億 | | 流通值 | 514.71億 | 流通股 | 2.45億 | | 市盈率TTM | 虧損 | 市盈率(靜) | 腐損 | | 市浮率 | 19.118 | 授手率 | 0.00% | | 婴比 | 94.74% | 量比 | | | 股息TTM | | 股息率TTM | | | 股息LFY | | 股息率LFY | | | 52週最高 | 214.000 | 52週最低 | 155.000 | | 歷史最高 | 214.000 | 歷史最低 | 155.000 | | 圆主 | 100股 | | | 作为专注通用 GPU 领域的硬核科技企业，天数智芯构建了覆盖 "天垓" 训练系列与 "智铠" 推理系列的完整产品矩阵，产品 兼容国内外主流 A ...

Group 1: Company Highlights - Three new stocks are set to be listed in Hong Kong on January 8, with all showing gains in the dark trading session prior to listing [1][2][3] - AI model developer Zhihui (2513.HK) closed at HKD 123.2, up 6.02% from the IPO price of HKD 116.2, indicating a profit of HKD 700 per lot of 100 shares [1] - Zhihui ranks first among independent general model developers in China and second among all general model developers, with a market share of 6.6% according to Frost & Sullivan [1] - Surgical robot company Jingfeng Medical-B (2675.HK) closed at HKD 60.15, a 39.11% increase from the IPO price of HKD 43.24, yielding a profit of HKD 1,691 per lot [2] - Jingfeng Medical focuses on designing, developing, and manufacturing surgical robots, with products in various stages of research and development [2] - General GPU chip and AI computing solution provider Tianshu Zhixin (9903.HK) closed at HKD 198.2, up 37.07% from the IPO price of HKD 144.6, resulting in a profit of HKD 5,360 per lot [2] - Tianshu Zhixin offers a range of general GPU products and AI computing solutions, targeting various industries including cloud computing and AI model development [2] Group 2: Market Performance - Zhihui's trading volume reached 3.5958 million shares, with a total market capitalization of HKD 54.236 billion [1] - Jingfeng Medical's trading volume was 1.7203 million shares, with a total market capitalization of HKD 23.321 billion [3] - Tianshu Zhixin's trading volume was 0.9513 million shares, with a total market capitalization of HKD 50.406 billion [3]

1. 美国三大股指期货涨跌不一，纳指期货跌0.19%，标普500指数期货跌0.05%，道指期货涨0.08%。 2. 欧股主要指数普跌，德国DAX指数涨0.63%，英国富时100指数跌0.63%，法国CAC指数跌0.05%，欧 洲斯托克50指数跌0.19%。 17. 自动驾驶技术公司Mobileye将以9亿美元收购机器人公司Mentee Robotics。 美股时段值得关注的事件： 21:15美国12月ADP就业人数 23:00 美国12月ISM非制造业PMI/美国11月JOLTs职位空缺/美国10月工厂订单月率 (格隆汇) 3. 美国总统特朗普：委内瑞拉的临时管理当局将向美国移交3000万至5000万桶高品质、受制裁的石油。 4. 白宫：特朗普及其团队正在讨论多种取得格陵兰岛的方案，包括从丹麦购买该领土、与格陵兰建立自 由联系协定，军事手段亦被列为选项之一。 5. 美国高院将1月9日定为今年首个意见发布日，特朗普关税案有望迎来裁决。 6. 美联储理事米兰：2026年需大幅降息超100基点，直言现行政策正在抑制经济。 7. 黄仁勋回应H200何时售往中国：正加快生产，等待许可的最终细节敲定。 8. AMD苏姿丰 ...

Core Viewpoint - The emergence of a new approach to semiconductor manufacturing, specifically the two-dimensional chip route, offers a potential breakthrough for China's chip industry, enabling the possibility of achieving 1nm technology domestically within five years without relying on EUV lithography [1][9]. Group 1: Two-Dimensional Chip Technology - Traditional silicon-based chip manufacturing relies on a "subtractive" process, which is complex and requires high precision [3]. - The two-dimensional semiconductor approach utilizes "additive manufacturing," allowing for the spontaneous and orderly growth of specific materials, significantly reducing manufacturing steps by 80% compared to traditional methods [3]. - This new method enhances performance in terms of leakage and heat control, addressing the power consumption challenges faced by AI and flagship mobile chips [3]. Group 2: Company Background and Achievements - The company behind this innovation, Yuanjiwei Technology, originated from a research team at Fudan University, founded in 2025 by Bao Wenzhong, who has extensive experience in two-dimensional semiconductor materials [5]. - Prior to the company's establishment, the team achieved significant milestones, including the development of the world's first two-dimensional chip, "Wuji," which integrated 5,900 transistors [5]. - Yuanjiwei has successfully launched China's first engineering verification demonstration line for two-dimensional semiconductors, marking a critical step from laboratory research to mass production [5]. Group 3: Production Roadmap - Yuanjiwei's ambitious roadmap includes completing equipment testing by June 2026, achieving small-scale production at 90nm by September 2026, and progressing to 28nm by 2027, with a target of reaching 1nm by 2030 [7]. - The phased approach, starting from established 90nm technology, aims to validate the stability and yield of the two-dimensional semiconductor manufacturing process [7]. - All equipment in this roadmap will utilize domestic supply chains, ensuring complete independence from foreign EUV technology [7]. Group 4: Industry Implications - The two-dimensional chip route presents an alternative to the traditional focus on EUV lithography, suggesting that China can redefine high-end chip manufacturing through innovation rather than following existing paradigms [9]. - This approach mirrors the success seen in the solar industry, where China became a leader by innovating rather than merely following established technologies [9]. - While challenges remain, such as material stability and yield improvement, the two-dimensional chip route signifies a critical shift in China's semiconductor strategy, potentially allowing it to transition from a follower to a leader in the global high-end chip market [9].

人工智能赛道蕴藏着巨大的市场机遇，投资者也普遍预期企业会为布局未来大举投入。 但如 今，英伟达公司（NVIDIA Corporation, NVDA）正将绝大部分现金用于收购与投资。这家 AI 芯片巨头近期不仅官宣了一项重磅人才招募计划，更敲定了一笔巨额收购交易 —— 种种动 向，都应被投资者视作 风险预警信号 。尽管如此，分析师对这只芯片股的投资主线观点仍偏 向 极度看多 ，只是市场必须警惕公司将超额现金用于激进投资可能引发的隐患。 收 购 G r o q 引 发 市 场 担 忧 在圣诞假期前后，英伟达抛出了一则重磅消息 —— 宣布收购一家虚拟技术初创企业。而在此 之前，这家 AI 图形处理器（GPU）巨头在过去一年里，就已频繁开展大规模投资与收购操 作。 从交易实质来看，英伟达斥资 200 亿美元，不仅获得了 Groq 公司的芯片技术授权，还收购 了该公司芯片业务相关的资产与核心团队。 值得关注的是，Groq 的首席执行官正是谷歌 （Alphabet, GOOG）张量处理单元（TPU）的初代开发者，且有消息称，Groq 在推理芯片 领域已取得重大突破。 【如需和我们交流可扫码添加进社群】 据一位前监管机构 ...

马云飞/摄 公开资料显示，作为国内最早布局AI加速领域的芯片企业，昆仑芯前身为百度智能芯片及架构部，由百度首席芯片架构师欧阳剑出任CEO，于2021年4 月完成独立融资，首轮融资估值即达130亿元。2025年7月，获得最新一轮股权融资，投后估值已达210亿元。IDC数据显示，2024年，中国GPU市场中，英 伟达以70%市场份额居首，华为昇腾以23%位列第二，昆仑芯排名第三。 海外投行Bernstein Research（伯恩斯坦）在2025年12月初发布了一份关于中国GPU市场份额的分析报告《China AI accelerators market share》。报告显 示，2025年华为以102.68亿美元销售额，占据国内AI加速器市场份额的40%，与英伟达的101.98亿美元基本持平，断层领先于其他国产GPU厂商。寒武纪以 9.99亿美元销售额与海光的11.18亿美元的销售额并列第三，阿里平头哥、百度昆仑芯、AMD等紧随其后。 对于拆分缘由，百度在上述公告中称，此举旨在提升昆仑芯在客户、供应商及潜在战略合作伙伴中的形象，并增强其在业务谈判中的地位，使百度集团 可通过持有的股份受益于昆仑芯的增长。此外，分 ...

个人观点仅供参考，文章中任何内容均不对您构成任何投资建议，据此操作，风险自担。 股市有风险，投资需谨慎。 压单又来了，就在今天收盘时，中信出现了14.5亿的压单，很多人担心，这是不是意味着行情要调整 了？其实不然，前面有两次类似的情况发生，分别是9月17日和10月24日，对应的压单有30亿和10亿， 30亿那次确实调整了一下，但是没两天就涨回来了，而10月24日之后是直接高举高打，所以这个压单顶 多就是心理层面的一些影响，又或者是上峰想要慢牛的一种体现，短期看来或许市场有调整的需求，但 是14连阳创历史的走法，意味着短期如果有调整，也会吸引更多资金涌入进来，因此调整大概率会成为 踏空者上车的好机会。 说回到题材上，真是铁打的商业航天，流水的其他题材，今天商业航天再次出现13家涨停，而且咱们说 的两个特征得到体现，一个是高位核心辨识度的高抛低吸，另一个是低位补涨挖掘，接下来还是这样， 直到这个方向出现亏钱效应再说。其次就是芯片和机器人，这个昨晚我们刚讲过，人形机器人和国产芯 片倒仍有一战之力，今天这两块都有表现，芯片更是涨停14家，比商业航天还强，不过别忘了前面说 的，铁打的商业航天，流水的其他题材，既然前有脑 ...

Core Viewpoint - The Ministry of Industry and Information Technology, along with eight other departments, has issued the "Implementation Opinions on the 'Artificial Intelligence + Manufacturing' Special Action" to enhance the supply of computing power for artificial intelligence [1] Group 1: Strengthening AI Computing Power Supply - The initiative emphasizes the development of intelligent chips through hardware-software collaboration, supporting breakthroughs in high-end training chips, edge inference chips, AI servers, high-speed interconnects, and intelligent cloud operating systems [1] - It aims to systematically advance the layout of high-level intelligent computing facilities and accelerate the construction of interconnected computing power platforms and a national integrated computing network monitoring and scheduling platform [1] - The plan includes pilot projects for intelligent cloud services and promotes the deployment of large model integrated machines, edge computing servers, and industrial cloud computing power to enhance the supply capacity of intelligent computing resources [1]

除了芯片进展之外，黄仁勋的本次演讲中重点讲到了物理AI（physical AI），并介绍了英伟达的Cosmos AI世界模型——开发者可使用Cosmos加速智能汽车、机器人和视频分析AI智能体的物理AI开发。 黄仁勋认为："物理AI的'ChatGPT时刻'近在咫尺，但挑战很明确。物理世界多样且不可预测。收集真实 世界训练数据缓慢又昂贵，而且永远不够。所以答案是合成数据。"他介绍，英伟达8年前开始研究自动 驾驶汽车，并透露："我们的愿景是有朝一日，每辆车都将是自主的，我们一直在为那个未来努力。" 据每日经济新闻报道，全球瞩目的2026国际消费电子展（CES）近日拉开帷幕。其中最受关注的，还是 美国英伟达公司创始人兼首席执行官黄仁勋的演讲。 这种生态锁定效应，让客户切换成本高达数千万美元，也让英伟达AI芯片毛利率远超行业平均水平。 所以，如果从这个角度来看，就会理解黄仁勋和英伟达为什么强调物理AI的价值。 现在，英伟达引以为傲的CUDA生态正成为物理AI落地的核心推手。黄仁勋强调的合成数据，其生成与 训练过程高度依赖并行计算能力。要模拟物理世界的千万种可能性，比如，从不同光照下物体的反光变 化到机械零件的散落状态 ...

Core Viewpoint - The Ministry of Industry and Information Technology, along with eight other departments, has issued the "Implementation Opinions on the Special Action of 'Artificial Intelligence + Manufacturing'" which emphasizes the strengthening of artificial intelligence computing power supply [1] Group 1: AI and Manufacturing Integration - The initiative aims to promote the collaborative development of intelligent chips and software/hardware [1] - Support is provided for breakthroughs in key core technologies such as high-end training chips, edge inference chips, AI servers, high-speed interconnects, and intelligent computing cloud operating systems [1] Group 2: Infrastructure Development - The plan includes an orderly advancement of high-level intelligent computing facility layouts [1] - It accelerates the construction of computing power interconnection platforms and a national integrated computing network monitoring and scheduling platform [1] Group 3: Service and Resource Enhancement - Pilot projects for intelligent computing cloud services will be launched [1] - The initiative aims to enhance the supply capacity of intelligent computing resources through the deployment of large model integrated machines, edge computing servers, and industrial cloud computing [1]