www.hczq.com 证券研究报告 | 医药生物 | 2026年01月31日 华创医药投资观点&研究专题周周谈 · 第161期 中国手术机器人行业近况更新 | 华创医药团队： | | | --- | --- | | 首席分析师郑辰 | 执业编号：S0360520110002邮箱：zhengchen@hcyjs.com | | 联席首席分析师刘浩 | 执业编号：S0360520120002邮箱：liuhao@hcyjs.com | | 医疗器械组组长李婵娟 | 执业编号：S0360520110004邮箱：lichanjuan@hcyjs.com | | 中药和流通组组长高初蕾 | 执业编号：S0360524070002邮箱：gaochulei@hcyjs.com | | 高级分析师王宏雨 | 执业编号：S0360523080006邮箱：wanghongyu@hcyjs.com | | 高级分析师朱珂琛 | 执业编号：S0360524070007邮箱：zhukechen@hcyjs.com | | 分析师陈俊威 | 执业编号：S0360525060002邮箱：chenjunwei@hcyjs.com | | ...

证券研究报告 | 医药生物 | 2026年01月31日 www.hczq.com 华创医药投资观点&研究专题周周谈 ·第161期 中国手术机器人行业近况更新 本周专题联系人：李焯娟 陈俊威 | 华创医药团队: | | | --- | --- | | 首席分析师郑辰 | 执业编号:S0360520110002邮箱:zhengchen@hcyjs.com | | 联席首席分析师刘浩 | 执业编号: S0360520120002邮箱: liuhao@hcyjs.com | | 医疗器械组组长李蝉娟 | 执业编号: S0360520110004邮箱: lichanjuan@hcyjs.com | | 中药和流通组组长高初营 | 执业编号:S0360524070002邮箱:gaochulei@hcyjs.com | | 高级分析师王宏雨 | 执业编号: S0360523080006邮箱: wanghongyu@hcyjs.com | | 高级分析师朱珂琛 | 执业编号: S0360524070007邮箱:zhukechen@hcyjs.com | | 分析师陈俊威 | 执业编号:S0360525060002邮箱:che ...

将基于四代架构陆续发布多款产品，持续提升计算性能。 本文为IPO早知道原创 作者｜ Stone Jin 微信公众号｜ipozaozhidao 据IPO早知道消息，天数智芯（股票代码：09903.HK）日前发布了四代架构路线图，提出以"高效 率、可预期、可持续"为核心的"高质量算力"设计目标，打造AI++算力系统新范式，预期于2027年 超越英伟达Rubin架构，迈向更具创新的突破性架构设计；发布边端算力产品"彤央"系列，实测性能 优于英伟达AGX Orin，并首次全面展示多行业深度应用案例及开放生态建设成果，以布局云边端的 全场景产品与解决方案，与产业链上下游的客户和生态伙伴共绘算力赋能千行百业的发展蓝图。 公布四代架构路线图 据天数智芯AI与加速计算技术负责人单天逸介绍： 2025年，天数天枢架构超越Hopper；2026年， 天数天璇架构对标Blackwell；2026年，天数天玑架构超越Blackwell；2027年，天数天权架构超 越Rubin；2027年之后将转向突破性计算芯片架构设计。 未来3年，天数智芯将基于此次发布的四 代架构，陆续发布多款产品，持续提升计算性能。 针对行业面临的能效比偏低、创 ...

Core Viewpoint - Arm has decided to develop its own chips, marking a significant shift from its traditional IP licensing model to a more direct involvement in chip manufacturing [1][3]. Group 1: Arm's Business Model and Market Position - Arm is known for its successful processor architecture, particularly in low-power, high-performance applications, widely used in billions of devices across mobile, embedded systems, and IoT [3][5]. - The company's IP licensing model allows various chip manufacturers to utilize its technology without fear of being "choked," fostering widespread adoption among companies like Xiaomi, MediaTek, and Apple [5][6]. - Arm's neutrality in the semiconductor field has been a key factor in its success, as it has acted as a technology provider without competing directly with its clients [10]. Group 2: Recent Financial Performance and Challenges - Arm's recent financial reports have shown troubling signs, with a 9% lower-than-expected revenue guidance following a record quarter of $1.24 billion in revenue and a 55% net profit growth [9]. - The company's net profit for the first quarter of fiscal year 2026 was $130 million, a 42% year-over-year decline, attributed to slowdowns in its core business areas: data centers, smart vehicles, and consumer electronics [9][11]. - Major clients like Tesla and Qualcomm are moving towards self-developed technologies, which poses a significant threat to Arm's traditional revenue model based on IP licensing [11]. Group 3: Strategic Shift and Market Implications - Arm's decision to enter chip manufacturing is seen as a response to declining revenues and the need to counteract clients who are attempting to bypass its licensing system [11][13]. - This move could lead to a major shake-up in the mobile chip market, potentially disrupting the current dominance of Qualcomm and MediaTek [13].

Core Viewpoint - Company is officially applying for a mainboard listing on the Hong Kong Stock Exchange, aiming to leverage AI technology for global expansion and product innovation [1] Group 1: Company Overview - Company is a leading global AI product firm, providing useful products to hundreds of millions of C-end users and diverse B-end enterprises [1] - Company has achieved significant revenue and net profit growth, with a projected compound annual growth rate (CAGR) of over 20% in revenue and a net profit margin exceeding 25% from 2022 to 2024 [1] Group 2: Technology and Innovation - Company has established three core technology platforms: "Tianshu," "Tianxuan," and "Tianji," which support its C-end and B-end product functionalities [3] - "Tianshu" provides distributed computing and storage for efficient AI functionality to millions of users across over 200 countries [3] - "Tianxuan" enhances machine understanding of language in complex scenarios, while "Tianji" aims to enable machines to think like humans [3] Group 3: Future Plans and Strategies - Post-listing, company will focus on advancing AI technology research, particularly in multi-modal large models and AGI goals [5] - Company plans to expand its global C-end product matrix and enhance B-end standardized products to cover more enterprise clients [5] - Company aims to seize strategic cooperation and investment opportunities while building a global talent system to support innovation and international growth [5]