致力于以芯片创新之力，赋能一个更加智能、更加美好的世界。 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，爱芯元智半导体股份有限公司（以下简称"爱芯元智"）于2026年2月10日正式 以"0600"为股票代码在港交所挂牌上市，成为"中国边缘AI芯片第一股"。 爱芯元智在本次IPO中发行104,915,200股H股。其中，香港公开发售获104.82 倍认购，国际发售 获6.8倍认购。同时，爱芯元智在本次发行中引入了一个多元化的基石投资阵容，累计认购1.85亿美 元。 截至10:35，爱芯元智报于每股28.90港元，市值170亿港元。 爱芯元智创始人、董事长兼执行 董事仇肖 莘博士在上市致辞中表示 ："此次上市是爱芯元智发展的 关键里程碑，也意味着更大的社会责任与新征程的开启。未来，我们将以平台型战略纵深布局，持续 突破技术边界、拓展业务版图，加速推动AI从智能感知向具身智能跃迁。 我们坚信，AI的价值主战场正在端侧与边缘侧全面展开。爱芯元智的边端AI芯片，正是这场变革的核 心基石。秉持'普惠AI，造就美好生活'的使命，践行'构建世界一流的感知与计算平 ...

Core Viewpoint - Aixin Yuanzhi, known as the first Chinese edge AI chip company, has completed its IPO and is now listed on the Hong Kong Stock Exchange under the stock code 00600.HK, with an issue price of 28.2 HKD and a total fundraising amount of 2.961 billion HKD [1]. Group 1: Company Overview - Aixin Yuanzhi opened flat at 28 HKD, with a market capitalization exceeding 16.59 billion HKD [2]. - The company was established in May 2019 and focuses on AI inference SoC chips, covering sectors such as terminal computing, edge computing, and smart vehicles. By September 2025, it is expected to have delivered over 165 million SoC chips and achieved five generations of SoC product development and commercialization [3]. Group 2: Financial and Shareholder Information - The net proceeds from the IPO will primarily be used for technology platform optimization, R&D investment, market expansion, and industrial investment mergers and acquisitions [4]. - The company is chaired by Dr. Qiu Xiaoxin, who leads strategic and technological innovation. CEO Sun Weifeng joined in April 2024, with a reported salary of 22.46 million CNY for the first nine months of 2025. Notable shareholders include Weichao Chuangxin and Qiming Venture Partners, each holding over 10%, along with investments from Tencent, Meituan, Lenovo Star, and Chongqing state-owned assets. "Chip Tycoon" Yu Renrong holds a 1.12% stake [4].

半导体行业的收并购浪潮，正从AI芯片厂商之间，进一步扩大到模拟芯片等多个领域。 仅在最近一周，就有德州仪器、英飞凌等厂商官宣新的收购动作，在业绩会上，高通则宣布上个季度完成了对Alphawave的收 购。 21世纪经济报道记者综合梳理发现，近期行业间发生的并购，已经从早期完善云端AI计算能力，进一步扩大到补全边缘AI落地 能力的阶段。 在当下这个AI应用加速之年，行业新一轮排位赛已悄然打响。 收并购加速 2月以来，多家模拟芯片巨头先后官宣了收购动作。 德州仪器宣布将以每股231美元的全现金交易方式收购Silicon Labs，旨在强化前者嵌入式无线连接解决方案的能力。 根据德州仪器分析，考虑到每天都有更多设备实现互联，此次交易将为公司新增约1200款支持多种无线连接标准和协议的产 品。 同时，德州仪器是IDM（垂直整合制造）模式厂商，旗下有自有晶圆生产制造厂，因此Silicon Labs此前的外部代工委托将完全 交给德州仪器完成，由此提高产品设计周期和上市速度。 该交易预计在2027年完成，德州仪器在公告中指出，此后三年内，该交易预计将产生约4.5亿美元的年度制造和运营协同效益。 记者查询Silicon La ...

此次发行引入16名基石投资者，合计认购1.85亿美元（约14.43亿港元）。其中韦尔半导体香港有限公司、新马服装国际有 限公司各认购3500万美元，JSC International Investment Fund SPC认购2700万美元，NGS Super Pty Limited认购2000万美 元，其余包括德赛西威、均胜电子旗下实体及多家机构投资者参与，形成产业资本与财务投资者结合的基石配置。 公司营收持续增长。2022年至2024年，营收从5023万元增至4.7亿元，三年复合增长率超过200%；2025年前九个月营收 2.69亿元，同比增加约5.8%。按2024年出货量计算，公司位居全球视觉端侧AI推理芯片供应商第五位，整体市占率6.8%， 中高端产品市占率24.1%；截至2025年9月末，累计出货量超过1.65亿颗，产品已覆盖视觉终端、智能汽车及边缘推理等多 个应用场景。 与营收增长并行的是持续亏损。公司2022年至2024年经营亏损分别为4.82亿元、5.94亿元、6.56亿元，经调整净亏损分别 为4.44亿元、5.42亿元、6.28亿元；2025年前九个月经营亏损5.49亿元，期内亏损8.56亿 ...

2026年被视作是AI原生应用大规模落地之年。无论创业类公司还是传统科技巨头，都在积极投身捕捉 新硬件形态的机遇并为此完善软件和生态闭环。 作为这些原生硬件的核心底座之一，嵌入式处理芯片由此也正面临新的发展机遇期：一方面，庞大的端 侧AI硬件正蓬勃出现，智能汽车、具身智能、AI手机等都是其中重要形态；另一方面，这些端侧AI也 面临高数据吞吐量和高性能、高集成度等需求，但仍要平衡成本和安全等新挑战。 近日，德州仪器（TI）嵌入式处理和DLP®产品高级副总裁Amichai Ron系统地阐述了公司面对端侧AI 浪潮的战略布局与实施路径。他强调，TI正从持续创新、产品可扩展性、产能保障三个层面构建应对体 系，以支持全球客户在AI时代实现技术落地。 作为嵌入式芯片领域的全球巨头，德州仪器旗下产品能覆盖端侧AI的传感、控制和处理等多个环节， 是支撑本轮AI大模型商业化应用的关键角色。 以自动驾驶技术为例，在CES 2026期间，TI发布了可扩展型TDA5高性能计算片上系统 (SoC) 系列，可 提供每秒10万亿次 (10 TOPS) 至每秒 1200 万亿次 (1200 TOPS) 运算的边缘人工智能算力，能满足L3级 ...

"整个世界都在快速变化，尤其是中国。"Ami c ha i坦言，包括他在内的TI人在密集与中 国市场创新者交流，以及时洞察并响应市场需求；同时TI也在加大研发投入，并加快产 品迭代速度，以更好把握这些新机遇。 边缘AI崛起 2026年被视作是AI原生应用大规模落地之年。无论创业类公司还是传统科技巨头，都在积极投身 捕捉新硬件形态的机遇并为此完善软件和生态闭环。 作为这些原生硬件的核心底座之一，嵌入式处理芯片由此也正面临新的发展机遇期：一 方面，庞大的端侧AI硬件正蓬勃出现，智能汽车、具身智能、AI手机等都是其中重要形 态；另一方面，这些端侧AI也面临高数据吞吐量和高性能、高集成度等需求，但仍要平 衡成本和安全等新挑战。 近日，德州仪器（TI）嵌入式处理和DLP®产品高级副总裁Ami c ha i Ron系统地阐述了 公 司 面 对 端 侧 AI 浪 潮 的 战 略 布 局 与 实 施 路 径 。 他 强 调 ， TI 正 从 持 续 创 新 、 产 品 可 扩 展 性、产能保障三个层面构建应对体系，以支持全球客户在AI时代实现技术落地。 作为嵌入式芯片领域的全球巨头，德州仪器旗下产品能覆盖端侧AI的传感、控制 ...

Core Insights - The report from Guolian Minsheng Securities emphasizes that the sustained high prosperity of AI upstream (computing power and storage) is a precursor to the current AI wave, while the market remains focused on the downstream applications of AI [1][2] - Clawdbot, representing edge AI, addresses market concerns regarding the diffusion of AI applications and the liberation of demand, indicating a significant potential for accelerating AI application expansion [2][3] AI Industry Trends - Since the beginning of the year, there have been numerous positive catalysts in the AI industry, with major tech companies like Meta and Tesla increasing their 2026 CapEx guidance, and AI's impact on advertising revenue becoming evident [1] - Recent price increases from cloud service providers reflect the ongoing strong demand for computing power and cloud resources under AI workloads, indicating a tight supply [1][2] Clawdbot's Impact - Clawdbot's open-source and pay-per-use business model, combined with its strong data privacy features and cost-efficiency, enables more non-large tech companies and individuals to access "self-owned AI," transforming AI from a high-end capability into a basic function for many devices [2] - The popularity of Clawdbot highlights the potential for accelerated AI application diffusion, although concerns about security vulnerabilities and exposure of sensitive data remain [2][3] Investment Recommendations - Companies to watch include Alibaba-W (09988), Baidu Group-SW (09888), Tencent Holdings (00700), Kingsoft Cloud (03896), and MiniMax-WP (00100), which provides Clawdbot's official API [4] - In the AI workflow sector, attention is drawn to Google (GOOGL.US), Microsoft (MSFT.US), Salesforce (CRM.US), Snowflake (SNOW.US), Elastic (ESTC.US), and MongoDB (MGDB.US) [4] - The demand for security driven by Clawdbot-like agents suggests monitoring Cloudflare (NET.US) and CrowdStrike (CRWD.US) [4]

IT之家 2 月 5 日消息，Arm 当地时间 4 日公布了该企业 2026 财年第 3 季度（2025 日历年第 4 季度）的财务数据。Arm 在上个季度实现 12.4 亿美元（IT之 家注：现汇率约合 86.18 亿元人民币）营收，连续四个季度位于十亿美元上方；同比增长 26%、环比增长约 9%。 Arm 的营收主要分为特许权、许可和其它 2 大板块，其中特许权收入同比增长 27% 至创纪录的 7.37 亿美元、许可和其它收入同比增长 25% 至 5.05 亿美 元。 Arm 上一财季签署了两份 CSS 许可协议，用于边缘 AI 平板电脑和智能手机；目前已有五家客户出货基于 Arm CSS 的芯片，其中两家出货的是其第二代平 台。 ...

Group 1 - Tencent's AI application Yuanbao launched a cash red envelope activity with a total scale of approximately 1 billion yuan, marking the largest consumer-end subsidy action in Tencent's AI sector to date [1] - On the same day, the Yuanbao App reached the number one position in the Apple Store's free app category, indicating strong user engagement [1] - Baidu's Wenxin Assistant red envelope activity has seen nearly 50 million users utilizing AI features, with photo search and AIGC video creation being the most popular [1] Group 2 - Domestic internet giants are heavily investing in the Spring Festival period to capture consumer-end AI traffic, signaling a new phase in AI application competition [2] - The Spring Festival is identified as a critical moment for the explosion of domestic AI applications due to its natural traffic advantages and increased user engagement during the holiday [2] - The competition among major AI model manufacturers is expected to accelerate user growth and commercialization, opening up growth opportunities in advertising, e-commerce, and entertainment [2] Group 3 - Tianrongxin has partnered with Tencent's Hongyuan model and is collaborating on various fronts including threat intelligence, model security, cloud security, privacy computing, and smart cities [3] - Yinghantong has simplified the deployment process of its edge computing machines, allowing for a budget-friendly setup that can operate continuously [3]

Core Insights - The global market is experiencing a dual storm of "AI leap" and "geopolitical restructuring" as it enters 2026, with Morgan Stanley predicting extreme volatility and Omdia forecasting a 31.26% growth in China's semiconductor market, reaching $546.5 billion [1] Group 1: Computing Power Dominance - The demand for computing power is expected to grow exponentially, driven by advancements in large language models (LLMs) and the adoption of "smart factory" models by LLM developers [1] - In China, the "edge AI era" is officially beginning, with digital terminals capable of edge inference (such as smartphones and vehicles) becoming the main drivers of semiconductor expansion [1] - Domestic AI chip suppliers in China are expected to capture a larger share of the market due to increased pressure from import bans [1] Group 2: Storage Chip Supercycle - The demand for high-performance storage (HBM) is surging as global AI infrastructure expands, leading to a significant upward revision of the storage market forecast by Omdia, which is now up by 62.8% [1] - China's reliance on Samsung and SK Hynix for 90% of its high-end storage results in low self-sufficiency and weak bargaining power, keeping average selling prices (ASP) of storage chips high [1] - A global shortage of storage chips is expected to persist until 2027-2028, which will be a major constraint on the shipment volumes of terminal devices [1] Group 3: Manufacturing Landscape in a Multipolar World - The U.S. is expected to see a resurgence in high-end manufacturing due to policies aimed at securing critical mineral and energy supplies, with automation technologies reducing the advantage of low-cost labor [1] - China aims to expand its market share in global tech manufacturing by leveraging its mature processes and vertical industry capabilities, despite facing external regulations [1] - In 2026, China is anticipated to achieve diversified breakthroughs in AI infrastructure through deep collaboration between computing power and AI ecosystems [1] Group 4: Energy and Deflationary Pressures - Energy costs are projected to become a critical factor in AI expansion, with AI giants moving towards "off-grid" strategies to control infrastructure directly [1] - By the second half of 2026, AI-driven cost reduction effects are expected to lead to rapid declines in the prices of certain goods, while the valuation of scarce assets that cannot be replicated by AI may rise [1]