公司的竞争力已不止于芯片设计。其打造的 "夸娥"（KUAE）智算集群已实现万卡级（10,240个GPU） 互联，具备支撑前沿大模型预训练的系统能力。这标志着摩尔线程完成了从硬件供应商向算力基础设施 平台的战略跃升，打开了更大的市场天花板。 光明的未来与现实的挑战 摩尔线程成立于2020年，创始团队成员均具有英伟达背景，创始人张建中曾任英伟达全球副总裁、大中 华区总经理。公司在短短四年内快速完成了四代GPU架构迭代，从"苏堤"、"春晓"到"曲院"、"平湖"， 产品覆盖 AI 智算、云计算和个人智算等高性能计算领域，产品出货形态包括、芯片、板卡、一体机、 集群。 财务数据展现出强劲的增长势头：营收从2022年的4600万元飙升至2024年的4.38亿元，三年复合增长率 超过200%，2025年上半年营收已达7.02亿元，接近去年全年的1.6倍。更引人注目的是，2024年至2025 年上半年，公司毛利率稳定在70%左右。显示出较强的产品竞争力。 然而，亮眼业绩的背后是持续的巨额研发投入。2022-2024 年，公司研发费用分别为 11.16、13.34、 13.59 亿元，为净亏损的主要原因。 本次发行定价充分体现 ...

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，摩尔线程智能科技（北京）股份有限公司（以下简称"摩尔线程"）今日正式开启 科创板IPO申购。 这意味着，A股或将迎来首家以"全功能GPU"为核心资产的上市公司。 A股或将迎来首家以"全功能GPU"为核心资产的上市公司。 台，能够支撑面向DeepSeek类前沿大模型预训练的万卡集群解决方案。这证明了摩尔线程已超越单 一的芯片设计，具备了提供大规模算力基础设施的平台能力。 而在生态建设上，摩尔线程推出"摩尔学院"开发者计划，并已完成超过百余家生态软件解决方案的 国产化兼容认证，其消费级显卡MTT S80在驱动持续更新后，已成功兼容近千款游戏和应用。这些 举措旨在推动成长飞轮的形成：更多的用户带来更丰富的应用场景反馈，驱动技术迭代，进而吸引更 多开发者，巩固生态优势。 成立于2020年的摩尔线程始终专注于全功能GPU的自主研发与设计，是国内高端AI芯片领域极为稀 缺的领军企业。 这里需要指出的一点是， 摩尔线程选择的"全功能GPU"路径，使其在国内厂商中展现出独特的战略 定位 ——与专注于AI计算的GPGP ...

市场或迎国产GPU第一股：成长与商业化是关键 11月24日，摩尔线程智能科技（北京）股份有限公司正式启动科创板IPO申购，A股或将迎来首家以"全 功能GPU"为核心资产的上市公司。 近年来，关于算力的话题几乎无处不在：大模型训练爆发、推理赶超、大规模智算中心建设落地、政策 密集出台、产业链全线提速——而GPU始终处于这一系列变革的核心位置。在AI算力日益成为全球战 略资源、科技竞争白热化的当下，资本市场终于迎来对国产GPU企业的首次系统性价值评估。 中国需要自己的英伟达，摩尔线程领跑全功能GPU 摩尔线程等一众AI芯片企业，自发展初期，便被市场赋予"中国版英伟达"的行业期待。 全球科技竞争格局正在重塑，高性能GPU算力，已成为数字经济时代的"石油"。然而，GPU领域技术壁 垒高、研发投入大、产业链条长，攻克难度极大。因此，无论是产业界还是资本市场，均需形成共识： 必须对这一领域的长期性与艰巨性有充分认知。 在此背景下，摩尔线程选择的"全功能GPU"路径，使其在国内厂商中展现出独特的战略定位。与专注于 AI计算的GPGPU或ASIC芯片不同，公司的核心产品是具备"功能完备性"与"精度完整性"的全功能 GPU。这 ...

今日，摩尔线程智能科技（北京）股份有限公司正式启动科创板IPO申购，A股即将迎来首家以"全功能GPU"为核心资 产的上市公司。 近年来，关于算力的话题几乎无处不在：大模型训练爆发、推理赶超、大规模智算中心建设落地、政策密集出台、产 业链全线提速——而GPU始终处于这一系列变革的核心位置。在AI算力日益成为全球战略资源、科技竞争白热化的当 下，资本市场终于要迎来了对国产"全功能GPU"企业的首次系统性价值评估。 然而，外界围绕摩尔线程的审视焦点始终不断：它究竟是国产GPU赛道中技术布局最全面的稀缺标的，还是一个仍需 时间验证的长期叙事？答案或许正藏在招股书的细节与数据之中。 一、中国需要自己的英伟达，摩尔线程领跑"全功能GPU" 摩尔线程等一众AI芯片企业，自发展初期，便被市场赋予"中国版英伟达"的行业期待。 全球科技竞争格局正在重塑，算力尤其是高性能GPU算力，已成为数字经济时代的"石油"。美国近年来连续升级对华 高端芯片出口管制，2023年10月将摩尔线程等国内头部AI芯片企业列入"实体清单"，直接限制英伟达H800、A800等高 性能AI芯片的对华出口。这一举措，恰恰从侧面反映出摩尔线程在技术实力和产业地 ...

Core Viewpoint - Moer Technology announced its initial public offering (IPO) and listing on the Sci-Tech Innovation Board, aiming to raise a total of 8 billion yuan through the issuance of 70 million shares at a price of 114.28 yuan per share [1][4]. Company Overview - The full name of the company is Moer Technology Intelligent Technology (Beijing) Co., Ltd., with the stock code 688795 and the abbreviation "Moer Technology" [2]. - The company focuses on the research, design, and sales of GPUs and related products, having launched four generations of GPU architectures since its establishment in 2020 [3][4]. IPO Details - The IPO will consist of 70 million shares, representing 14.89% of the total share capital post-issuance [2]. - The expected total fundraising amount is 8 billion yuan, with a net amount of approximately 7.576 billion yuan after deducting issuance costs [1][2]. - The pricing method for the issuance is based on preliminary inquiries, with no further bidding for the offline portion [2]. Product Line and Market Focus - Moer Technology's product line includes AI training and inference cards, graphic acceleration products for high-end applications, and consumer-grade graphic acceleration products for AI PCs and gaming PCs [4]. - The company aims to provide integrated solutions across cloud, edge, and terminal markets, enhancing its capabilities in high-performance computing [4]. Use of Proceeds - The funds raised from the IPO will be allocated to the development of next-generation AI training and inference chips, graphic chips, and AISoC chips, as well as to supplement working capital [4].

Investment Rating - The report assigns a rating based on the AHP score, with a score of 3.39 indicating a position in the upper-middle tier of the AHP model for the Sci-Tech Innovation Board [8][11]. Core Insights - The company focuses on the research, design, and sales of self-developed full-function GPUs, with a strategic shift towards high-margin, high-performance segments, particularly in AI computing [11][19]. - The company has achieved significant advancements in GPU architecture, enabling breakthroughs in computing power and product performance that meet international standards [19][20]. - The AI chip market in China is projected to reach 1.3368 trillion yuan by 2029, with a compound annual growth rate of 53.7% from 2025 to 2029, indicating substantial growth potential for the company [19][24]. Summary by Sections 1. AHP Score and Expected Allocation Ratio - The AHP score for the company is calculated at 3.39, placing it in the 56.1% percentile of the AHP model, indicating a strong position in the market [8][10]. 2. Highlights and Features of the New Stock 2.1 Core Team and Self-Developed GPU Architecture - The core team comprises professionals with extensive experience from leading international companies, focusing on self-developed full-function GPUs [11][15]. 2.2 Full-Function GPU Achievements - The company has developed a comprehensive product system that includes AI computing, cloud computing, and personal computing applications, with significant revenue growth expected from AI computing products [11][19]. 2.3 Capitalizing on AI and Graphics Opportunities - The company plans to invest in the development of next-generation AI training and inference chips, graphics chips, and AI SoC chips, aiming to strengthen its product performance and technology barriers [24][26]. 3. Comparable Company Financial Metrics 3.1 Revenue and Profit Trends - The company’s revenue is projected to grow significantly, with estimates of 0.46 billion yuan in 2022, increasing to 7.02 billion yuan by the first half of 2025, while net losses are expected to narrow [28][30]. 3.2 R&D Expenditure Trends - R&D expenses have been increasing, with the company spending 11.16 million yuan in 2022 and 5.57 million yuan in the first half of 2025, indicating a strong commitment to innovation [34][36]. 3.3 Profitability and Debt Levels - The company’s gross margin has improved significantly, with a gross margin of 70.71% in 2024, while the debt-to-asset ratio has been relatively high, indicating potential financial risks [39][40]. 4. Fundraising Projects and Development Vision - The company plans to raise funds for the development of new AI training and inference chips, graphics chips, and AI SoC chips, which are expected to enhance its competitive position in the GPU market [42][43].

Market Overview - The market experienced wide fluctuations throughout the day, with the ChiNext index dropping over 2% before rebounding strongly in the last trading hour [1] - In the afternoon, the market narrowed its trading volume, but funds collectively targeted hard technology sectors, leading to a recovery in indices and multiple core stocks reaching historical highs [1] Semiconductor Sector Performance - Chip stocks showed localized activity, with previously lagging sectors like silicon wafers and equipment seeing significant gains [1] - Longchuan Technology announced a substantial profit increase of 20% for the first three quarters, leading to a trading halt, while other companies like Jingzhida and Hushil Industry rose over 10% [1][6] - The storage chip segment maintained strong popularity, with Demingli achieving three consecutive trading limits and Jucheng shares rising over 10% [1] ETF Performance - Semiconductor equipment ETFs rose over 7%, semiconductor materials ETFs increased by over 6%, and various semiconductor device ETFs saw gains exceeding 5% [1][3] Positive News Catalysts - Nvidia and OpenAI announced a partnership involving a massive data center plan and Nvidia's potential investment of up to $100 billion in OpenAI, marking a significant step in their decade-long collaboration [5] - Longchuan Technology projected a net profit of 827 million to 877 million yuan for the first three quarters, representing a year-on-year increase of 131.39% to 145.38% due to strong market demand in the semiconductor industry [6] - The World Semiconductor Trade Statistics Organization (WSTS) forecasts global semiconductor market sales to reach $700.9 billion in 2025, with an expected growth of 11.2% [7]

Group 1: Stock Market Activity - On the institutional trading leaderboard, 32 stocks were listed, with 16 showing net buying and 16 showing net selling [1] - The top three stocks with the highest net buying by institutions were ZTE Corporation (1.105 billion), Kunlun Wanwei (274 million), and Heertai (262 million) [1] - ZTE Corporation saw a price increase of 9.38%, Kunlun Wanwei increased by 20.01%, and Heertai rose by 10.02% [1] Group 2: Domestic Chip Industry - Domestic AI chip manufacturers like Cambricon and Haiguang Information are gaining traction, with Haiguang planning to merge with Zhongke Shuguang [2] - The domestic GPU market is entering a rapid development phase, transitioning from "usable" to "good" [2] - Companies like Moer Thread are enhancing AI computing power with their latest GPU architecture, supporting advanced AI model training [2] Group 3: GPU Market Forecast - According to Frost & Sullivan, the global GPU market is expected to reach 3.611974 trillion yuan by 2029, with China's GPU market projected to reach 1.363578 trillion yuan [3] - China's market share in the global GPU market is anticipated to increase from 30.8% in 2024 to 37.8% by 2029 [3] Group 4: RWA and Digital Currency - Zhongyou Capital's subsidiary, Kunlun Bank, is the only official bank for international settlements between China and Iran, and is involved in digital RMB payment services [4] - Yuyin Co. has established a digital currency research center to explore applications in digital currency transactions [4] - RWA is seen as a means to enhance efficiency and reduce costs, facilitating financial globalization and decentralization [4]

Group 1 - The STAR Market chip index showed strong performance with an intraday increase of 7.33%, and the related ETF, the STAR Chip ETF (588290), surged by 7.15% with a trading volume exceeding 200 million yuan, reflecting a year-to-date increase of over 26% [1] - Key component stocks such as Shengmei Shanghai hit the daily limit, while Haiguang Information and Cambricon Technologies rose over 10% [1] - The DeepSeek-V3.1 was officially released, featuring significant adjustments in tokenizer and chat templates, and is designed for the upcoming generation of domestic chips [1] Group 2 - Zhongyuan Securities predicts that by the first half of 2025, technological innovation will drive the rise of domestic large models, with widespread AI applications and the launch of AI glasses, marking the beginning of an era of universal intelligent driving [2] - The semiconductor industry is expected to continue its recovery, with increasing demand for self-controlled solutions and robust growth in AI computing power on the cloud side [2] - The storage market is rebounding, with customized storage becoming a trend, enhancing the competitive advantage of domestic storage module manufacturers and accelerating the growth of domestic production [2] Group 3 - Huaxin Securities emphasizes the necessity of domestic high-end AI chip localization, with companies like Cambricon, Haiguang Information, and Huawei leading the charge, while other domestic AI chip manufacturers are also accelerating their market penetration [3] - Moore Threads provides AI training and inference cards, with its latest GPU architecture significantly enhancing AI computing power, supporting advanced model pre-training solutions [3] - Muxi offers a series of GPUs with strong parallel computing capabilities, already applied at scale in multiple intelligent computing centers, and also provides inference-focused GPUs for cloud AI scenarios [3]

Group 1 - DeepSeek's latest V3.1 version utilizes UE8M0 FP8 Scale parameter precision, designed for the upcoming domestic chip release [1] - FP8 is a cutting-edge low-precision format for AI computing, significantly enhancing GPU performance and reducing memory usage for large language model training [1] - Domestic GPUs are rapidly developing, transitioning from "usable" to "user-friendly" stages, although they have not yet matched international products [1] Group 2 - Companies like Cambrian, Haiguang Information, and Huawei are leading the A-share computing chip market [3] - Moer Thread provides AI training and inference cards, with its latest GPU supporting FP8 precision, significantly boosting AI computing power [1] - Muxi offers C series GPUs for integrated training and inference, and N series GPUs focused on cloud AI inference, showcasing strong mixed-precision computing capabilities [2] Group 3 - The global GPU market is projected to reach 36,119.74 billion yuan by 2029, with China's GPU market expected to reach 13,635.78 billion yuan, increasing its global market share from 30.8% in 2024 to 37.8% in 2029 [2] - DeepSeek is driving the shift of AI applications from centralized cloud services to mass terminals, necessitating high-cost performance dedicated chips [2] - The domestic chip manufacturers and application enterprises are accelerating their integration with DeepSeek, anticipating a significant increase in domestic computing power by 2025 [2] Group 4 - Huawei's Ascend ecosystem includes companies like Tuo Wei Information, Digital China, and Huafeng Technology, enhancing mixed inference architecture and agent capabilities [4] - The upgraded model shows significant improvements in tool usage and intelligent agent tasks through Post-Training optimization [4] Group 5 - Related companies include Dingjie Zhizhi, Fanwei Network, and Kute Intelligent [5]