11月24日，有"中国版英伟达"之称的"国产GPU第一股"的摩尔线程正式启动科创板发行，发行价定为114.28元/股，创下2025年以来A股新股发行价新高。 11月24日，摩尔线程公告，根据上交所提供的数据，本次网上发行有效申购户数为482.66万户，有效申购股数为462.17亿股，网上发行初步中签率为 0.02423369%。根据《摩尔线程智能科技（北京）股份有限公司首次公开发行股票并在科创板上市发行公告》公布的回拨机制，由于本次网上发行初步有效 申购倍数约为4126.49倍，超过100倍，发行人和保荐人（主承销商）决定启动回拨机制，对网下、网上发行的规模进行调节，将扣除最终战略配售部分后本 次公开发行股票数量的10%（向上取整至500股的整数倍，即560.00万股）从网下回拨到网上。 回拨机制启动后：网下最终发行数量为3,920.00万股，占扣除最终战略配售数量后发行数量的70.00%；网上最终发行数量为1,680.00万股，占扣除最终战略 配售数量后发行数量的30.00%。回拨机制启动后，网上发行最终中签率为0.03635054%。 11月20日晚间，摩尔线程发布了《首次公开发行股票并在科创板上市发行公 ...

Core Viewpoint - Two new stocks are available for subscription next week, with Moer Thread set to become the "first domestic GPU stock" in China [2] Group 1: Moer Thread - Moer Thread's subscription code is 787795, with an issue price of 114.28 yuan per share and a total issuance of 70 million shares, of which 11.2 million shares are for online subscription [2] - The company focuses on the research, design, and sales of GPUs and related products, having launched four generations of GPU architecture since its establishment in 2020 [2] - Moer Thread's projected revenues for 2025 are between 1.218 billion yuan and 1.498 billion yuan, representing a year-on-year growth of 177.79% to 241.65% [5][6] - The net profit attributable to the parent company for 2025 is expected to be between -1.168 billion yuan and -730 million yuan, showing a year-on-year improvement of 27.82% to 54.89% [5][6] Group 2: Bai Ao Sai Tu - Bai Ao Sai Tu's subscription code is 787796, with the issue price and P/E ratio yet to be disclosed, but the industry P/E ratio is referenced at 39.58 times [2] - The company operates in the preclinical CRO and biotechnology sectors, with four technology platforms, including gene editing and antibody drug discovery [7] - Bai Ao Sai Tu's projected revenue for 2025 is 1.351 billion yuan, reflecting a year-on-year growth of 37.75%, with a net profit attributable to the parent company expected to be 135.37 million yuan, a significant increase of 303.57% [10]

根据目前发行安排，下周有2只新股可申购。 具体来看，11月24日可申购上交所科创板新股摩尔线程，11月28日可申购上交所科创板新股百奥赛图。 摩尔线程将成"国产GPU第一股" 【导读】下周2只新股可申购，摩尔线程将成"国产GPU第一股" 中国基金报记者 闻言 A股打新投资者注意啦！ 摩尔线程的申购代码是787795，发行价为114.28元/股，此次发行总数为7000万股，其中网上发行数量为1120万股。 投资者参与网上申购摩尔线程，申购数量上限为1.1万股，顶格申购需配沪市市值11万元。 摩尔线程主要从事GPU（图形处理器）及相关产品的研发、设计和销售，自2020年成立以来已推出四代GPU架构，并拓展出覆盖AI智算、 云计算和个人智算等应用领域的计算加速产品矩阵。 在高端通用芯片领域，GPU构成了电子计算设备的智慧核心与运算枢纽，对推动AI、大数据、云计算等前沿科技的进步起到决定性作用， 但GPU的自主研发面临的技术壁垒极高。 目前，英伟达、AMD（超威半导体）等企业 在 全球集成电路 市场 占据主导地位，摩尔线程等中国企业起步相对较晚，但已在各自专攻领 域获取 了 一定的市场份额，推动 了 中国GPU产业的 ...

Core Viewpoint - Nvidia's impressive earnings report did not lead to the expected market rally, instead resulting in a significant market downturn, highlighting the sensitivity of high-valuation stocks to interest rate expectations [1][5][11]. Financial Performance - Nvidia reported Q3 2026 fiscal year revenue of $57 billion, a year-over-year increase of approximately 62% and a quarter-over-quarter increase of about 22% [2]. - The data center business generated $51.2 billion in revenue, reflecting a year-over-year growth of 66% and a quarter-over-quarter growth of 25% [2]. - The GAAP gross margin was around 73%, indicating strong profitability [2]. - Nvidia's management provided a Q4 revenue guidance of approximately $65 billion, significantly above market expectations [2]. Market Reaction - Following the earnings report, futures and after-hours stock prices rose, with the S&P 500 and Nasdaq initially showing strong gains [3]. - However, the market reversed course within half a day, leading to a decline in major indices and Nvidia's stock price [4][5]. Economic Indicators - The release of the delayed September U.S. non-farm payroll report showed mixed results, with job additions of approximately 119,000, exceeding expectations of 50,000, but the unemployment rate rose to 4.4% [6][7]. - This data suggested a robust economy, which tempered expectations for immediate interest rate cuts by the Federal Reserve [7]. Technical Factors - A significant amount of options expired, triggering mechanical selling pressure in the market [9]. - The market experienced a broader risk reduction and deleveraging, affecting not only U.S. stocks but also cryptocurrencies [10]. Investment Insights - The day’s events indicated a re-evaluation of valuations and positions rather than a fundamental decline in the AI sector [11]. - Investors are advised to maintain a balanced approach to AI investments, diversifying across various assets rather than concentrating on a single stock [14].

今年 3 月，我曾向大家提及包括 CoreWeave 在内的部分人工智能公司面临的折旧风险。8 月，做 空安然公司的吉姆・查诺斯也表达了类似担忧。 核心顾虑集中在 GPU（图形处理器）上，这类芯片是训练和运行人工智能模型的关键硬件。随着 新型 GPU 推出，旧款产品会因技术淘汰和磨损而贬值。云计算公司必须通过折旧方式，在符合实 际的周期内降低这些资产的账面价值，折旧速度越快，对盈利的冲击就越大。 投资者开始担心，GPU 的实际使用寿命仅为 1 到 2 年，但云计算提供商却按 5 到 6 年对其进行 折旧。这种会计处理上的不匹配，可能会导致人工智能行业在几年后遭遇严重的盈利下滑。 如果您希望可以时常见面，欢迎标星收藏哦~ 来 源 ： 内容编译自 businessinsider ， 谢谢 。 近几周，关于人工智能泡沫的悲观警告震动了市场，但至少有一项主要担忧是站不住脚的。 如今，这种观点在华尔街几乎已成共识，也是支撑 "我们正处于巨大人工智能泡沫中" 这一论点的 主要依据之一。但问题在于，这种看法是错误的：即便英伟达每 18 个月或更短时间就推出新一代 GPU 架构，其折旧速度也远没有部分投资者担忧的那么快。 " ...

Group 1: Market Overview - The semiconductor chip market experienced a rise followed by a decline, with the first Hong Kong ETF focused on the "Hong Kong chip" industry chain (159131) dropping by 0.51% in price, while trading volume exceeded 31 million CNY, indicating active trading [1]. Group 2: Huawei's AI Technology - Huawei is set to release a groundbreaking AI technology on November 21, which aims to improve the utilization efficiency of computing resources from the industry average of 30%-40% to 70%, significantly enhancing the potential of computing hardware [3]. - This technology will unify resource management across various computing hardware, including GPUs and NPUs, providing more efficient support for AI training and inference [3]. - The technology shares similarities with the core technology of Israeli AI startup Run:ai, which was acquired by NVIDIA for 700 million USD at the end of 2024 [3]. Group 3: Domestic AI Chip Market - According to CITIC Securities, the domestic AI chip market is entering a high growth phase, with accelerated commercialization of applications [3]. - There is a trend of increasing orders for domestic chips as cloud vendors and AI startups shift their procurement plans towards domestic chips due to stable supply demands [3]. - The rapid growth in domestic computing consumption is expected to further accelerate the release of domestic AI chips [3]. Group 4: Hong Kong Chip ETF Details - The first Hong Kong ETF focusing on the "Hong Kong chip" industry chain has a composition of 70% hardware and 30% software, heavily investing in semiconductor, electronics, and computer software sectors, with notable weights for companies like SMIC (20.27%), Xiaomi Group-W (9.11%), and Hua Hong Semiconductor (5.64%) [4]. - This ETF excludes major internet companies like Alibaba, Tencent, and Meituan, allowing for a sharper focus on the AI hard technology sector in Hong Kong [4].

消息面，华为将在11月21日发布AI领域的突破性技术，有望解决算力资源利用效率的难题。具体 来看，华为即将发布AI领域的突破性技术，可将GPU（图形处理器）、NPU（神经网络处理器）等算 力资源的利用率，从行业平均的30%至40%提升至70%，显著释放算力硬件潜能。 注：如提及个股仅供参考，不代表投资建议。指数/基金短期涨跌幅及历史表现仅供分析参考，不 预示未来表现。市场观点随市场环境变化而变动，不构成任何投资建议或承诺。文中提及指数仅供参 考，不构成任何投资建议，也不构成对基金业绩的预测和保证。如需购买相关基金产品，请选择与风险 等级相匹配的产品。基金有风险，投资需谨慎。 （责任编辑：董萍萍 ） 国信证券表示，AI对互联网巨头广告业务场景、云计算场景和企业效率方面的作用依然明显，各 公司依然保持积极的投资力度，但市场开始逐渐增加关注巨额CAPEX背后的ROI表现。国内大厂的资本 开支压力远小于海外，且AI对其相关业务的带动同样明显，同时利润端预计多家龙头公司经营效率提 升明显。 【免责声明】本文仅代表作者本人观点，与和讯网无关。和讯网站对文中陈述、观点判断保持中立，不对所包含内容 的准确性、可靠性或完整性提 ...

Group 1 - AIGC, AI computing power, and ChatGPT sectors experienced a strong rebound, with stocks like Dongfang Guoxin and BlueFocus rising over 10% [1] - The AI sector is seeing increased capital inflow, with the Huaxia AI ETF (159381) rising 1.82% and the 5G Communication ETF (515050) up 0.61% [1] - Huawei is set to release breakthrough technology in the AI field, potentially increasing the utilization rate of GPU and NPU resources from the industry average of 30%-40% to 70% [1] Group 2 - Bohai Securities noted that the demand side for AI computing power is driven by continued high capital expenditure from cloud vendors and a growing consensus on domestic computing power [1] - Dongguan Securities highlighted that tech giants like Inspur and Huawei are actively developing supernode technology, aiming to create a self-controlled and open domestic computing foundation [2] Group 3 - The Huaxia AI ETF (159381) tracks the ChiNext AI Index and has a significant allocation to optical modules, with the top three holdings being Zhongji Xuchuang, Xinyi Sheng, and Tianfu Communication [3] - The 5G Communication ETF (515050) focuses on the 5G communication theme index and has a scale exceeding 9 billion, with major holdings including Zhongji Xuchuang, Xinyi Sheng, and Luxshare Precision [3]

Group 1 - Huawei is set to release a breakthrough technology in the AI field on November 21, which aims to improve the efficiency of computing resource utilization from the industry average of 30%-40% to 70% [1] - The new technology will enable unified resource management and utilization of computing power from various sources, including GPUs and NPUs, enhancing support for AI training and inference [1] - Huawei's AI chip roadmap includes three series of products: 950PR/950DT, 960, and 970, with the 950PR expected to launch in Q1 next year, featuring enhanced interconnect bandwidth and computing power [1] Group 2 - Analysts from Dongfang Securities believe that the future of computing power competition will shift from GPUs to supernodes, which will become the mainstream form of computing power [2] - The emergence of large-scale supernodes, such as those showcased by Nvidia and Huawei, is expected to significantly enhance the training efficiency of domestic large models in China [2] - The domestic computing power market is anticipated to experience a boom in both large model training and AI capital expenditure next year [2] Group 3 - Tuowei Information is a strategic partner of Huawei in the "Cloud + Kunpeng/Ai + Industry Large Model + Open Source Harmony" field, developing innovative solutions based on the Harmony operating system [3] - Huafeng Technology, as a high-speed line module supplier, has established deep partnerships with leading domestic AI server manufacturers, including Huawei, ZTE, and Alibaba [3]

IT之家 11 月 11 日消息，别再纠结实验室面积与人员编制了，现代科研机构的真正"硬通货"是 GPU（图形处理器）。至少，马克・扎克伯格（Mark Zuckerberg）与普莉希拉・陈（Priscilla Chan）是这么认为的。 这对夫妇表示，其慈善机构"陈-扎克伯格倡议"（Chan Zuckerberg Initiative, CZI）当前最迫切的需求是算力，而非其他资源，以支撑其最新确立的、以科 学与人工智能为核心的战略使命。 Biohub 在公告中明确表示，计划到 2028 年将其算力集群扩展至 10,000 块 GPU。陈在播客中透露，CZI 目前 GPU 集群规模已达 1,000 块，并重申了迈 向"万卡（10,000 GPU）量级"的明确规划。她此前亦曾表示，CZI 旨在以强大的算力资源吸引顶尖生物科研人才，而非试图以媲美科技巨头的薪资待遇 进行竞争。 在 11 月 6 日播出的《a16z 播客》节目中，陈在谈及 CZI 自 2016 年起支持的生物学实验室集群"Biohub"时表示："我们并未大规模扩建物理空间，但我 们正在显著扩充算力。" 据IT之家了解，CZI 于 2015 年由扎克伯 ...