◎记者 严晓菲 这种看似矛盾的现象，源于各方对国产GPU市场广阔前景的共识。据国际数据公司（IDC）预测，到 2029年，中国加速服务器市场规模将超过千亿美元。交银国际研报认为，国产算力芯片行业未来将呈现 百花齐放态势，技术上取得长足进步，包括晶圆代工、存储器、半导体设备等在内的国产算力产业链有 望持续受益。 12月以来，国产GPU领域的两位主要玩家——摩尔线程和沐曦股份，先后叩开了资本市场的大门。这两 家成立均仅5年的公司，正以超预期的市场表现，点燃投资者的巨大热情。 沐曦股份上市首日便坐上A股史上"最赚钱新股"的宝座，股价一度飙涨到逼近900元/股，总市值也已突 破3300亿元。在此之前，摩尔线程在上市后短短几个交易日内，股价最高涨到941.08元/股。 在这两家公司的背后，是两份密密麻麻的出资方名单：沐曦股份股东超过120家，摩尔线程股东超过80 家。在这场从"非共识"到"共识"的转变中，这些"耐心资本"为何能如此守得住寂寞？ 高光时刻： 亏损与高估值反差折射广阔前景 在国产GPU的高光时刻，难以无视的是这两家公司依旧严峻的财务现实。根据招股书，它们都仍处于持 续亏损的状态。 沐曦股份2022至2024 ...

12月10日，摩尔线程再度成为市场焦点。盘中一度涨超26%，最高触及797.97元/股，总市值超3500亿 元，创上市新高。截至发稿，涨幅达16%，报733.22元/股，成为A股市场第三高价股，仅次于寒武纪和 贵州茅台。 实际上，作为科创板年内最大IPO，摩尔线程自启动上市便备受市场期待。其网上发行阶段共吸引 482.66万户投资者参与抢筹，最终中签率仅0.036%。发行价定为114.28元/股，上市首日开盘即暴涨 468.78%，中一签盈利近27万元，成为年内最赚钱新股。 此次股价创新高的催化剂或是公司即将举行的首届MUSA开发者大会。 据其官微消息，摩尔线程创始人张建中将在12月19日—20日举办的开发者大会上，首次系统阐述以 MUSA为核心的全栈发展战略与未来愿景，并发布新一代GPU架构、推出涵盖产品体系、核心技术及 行业解决方案的完整布局。 行业层面，GPU市场正迎来爆发式增长。据沙利文预测，中国GPU市场规模将从2024年的1425亿元跃升 至2029年的13368亿元，2025-2029年CAGR高达53.7%，增速远超全球市场。国金证券也在研报中指 出，AI时代驱动GPU需求快速扩张，国产化进入 ...

Core Viewpoint - The article highlights the remarkable debut of Moer Thread, the first domestic GPU company, on the STAR Market, with its stock price soaring by 468.78% on the opening day, reflecting strong market enthusiasm for domestic high-end computing chips amid the AI boom [3][4]. Group 1: Company Overview - Moer Thread was founded by Zhang Jianzhong, a veteran in the GPU industry, who previously worked at NVIDIA for 14 years, establishing a complete GPU ecosystem in China [6][8]. - The company has rapidly evolved from a startup to a leading domestic GPU manufacturer within three years, launching its first full-function GPU, Sudi, in just 19 months [8][10]. Group 2: Financial Performance - Moer Thread's revenue structure has shifted significantly, with AI computing products contributing 77.63% of total revenue by 2024, up from 71.44% from desktop graphics acceleration products in 2022 [11]. - The company's revenue grew explosively from 0.46 billion yuan in 2022 to 4.38 billion yuan in 2024, achieving a compound annual growth rate of 208.44% [11]. - Despite cumulative losses exceeding 5 billion yuan from 2022 to 2024, the loss margin is narrowing, with expectations of profitability by 2027 [11]. Group 3: Capital Market Activity - Moer Thread achieved a valuation of over 1 billion USD within 100 days of its establishment, setting a record for the fastest unicorn ascent in the industry [12][15]. - The company has raised over 9 billion yuan in funding from 2021 to 2024, with a valuation reaching around 30 billion yuan [12][15]. Group 4: Future Prospects - The IPO raised approximately 7.576 billion yuan, which will be invested in three core technology projects, including the development of AI training chips and graphics chips [18]. - The global AI chip market is projected to reach 3.62 trillion yuan by 2029, with China's market expected to grow at a compound annual growth rate of 51.1% [20]. - The domestic GPU sector is entering a golden era, with other companies like Muxi and Biran Technology also planning to go public, indicating a competitive landscape [20].

Core Points - Moer Technology, known as the "first domestic GPU stock," officially listed on the Sci-Tech Innovation Board on December 5, 2025, with an opening price of 650 CNY per share, marking a 468.78% increase from its IPO price of 114.28 CNY [1][3] - The company aims to raise 8 billion CNY through this IPO, primarily for the development of next-generation AI training and inference chips, as well as other GPU projects [7][10] - Moer Technology has raised over 10 billion CNY in total funding since its inception, with a notable investor base including Sequoia Capital and Tencent [8][10] Company Overview - Moer Technology was founded in June 2020 by Zhang Jianzhong, a former vice president of NVIDIA, and has quickly gained attention in the GPU market [8][10] - The company has launched four generations of GPU architecture chips from 2021 to 2024, with plans to release an intelligent SoC chip named "Changjiang" in 2024 [7][10] - As of the latest financial reports, Moer Technology has not yet achieved profitability, with projected revenues of 0.46 billion CNY, 1.24 billion CNY, and 4.38 billion CNY for 2022 to 2024, respectively [10][12] Market Position - Moer Technology's full-function GPU is positioned to compete directly with NVIDIA, offering unique advantages in AI training, rendering, and high-performance computing [15] - The company has a low subscription rate of approximately 0.0363% for its IPO, indicating high demand but limited access for individual investors [5][10] - The competitive landscape is intensifying, with other domestic companies like Muxi Technology also pursuing IPOs, which could impact Moer Technology's market share [14][15]

Core Viewpoint - The capital market's unprecedented enthusiasm for computing power concepts is exemplified by the explosive debut of Moole Thread on the STAR Market, with a first-day opening price of 650 CNY per share, leading to a market capitalization of 305.5 billion CNY, marking a 468.78% increase from its issuance price [1][5] Company Overview - Moole Thread is often referred to as the "Chinese version of NVIDIA" and is recognized as the "first stock of general-purpose GPU in China," attracting significant market speculation [1][5] - The company has accumulated losses of nearly 6 billion CNY over three years and relies heavily on external funding [1][5] Market Positioning - The core logic behind market optimism for Moole Thread lies in its strategic rarity in the domestic general-purpose computing sector, which is a key reason for its comparison to NVIDIA [1][5] - The company has a full-stack self-research technology route, with capabilities from architecture definition to packaging and testing, and has developed a series of GPU architectures [2][5] Product Development - Moole Thread has launched four generations of GPU architectures, with the latest "Pinghu" achieving an FP32 Vector computing power of 32T, surpassing NVIDIA's A100 but below the H100 [2][5] - The company has made significant breakthroughs in cutting-edge technologies such as FP8 efficient computing, positioning itself among the few domestic manufacturers with such capabilities [2][5] Business Model - The company adopts a Fabless model, focusing on the research, design, and sales of GPU chips while outsourcing manufacturing and packaging to specialized firms, optimizing operational and capital efficiency [3][5] Financial Performance - Moole Thread has shown rapid revenue growth, with projected revenues of 0.46 billion CNY in 2022, 1.24 billion CNY in 2023, and expected revenues of 4.38 billion CNY and 7.85 billion CNY in the first three quarters of 2024 [6][8] - The company’s gross margin has improved significantly, from -70% in 2022 to 61.9% in the first three quarters of 2025, indicating the emergence of scale effects [9][10] Market Potential - The AI chip market in China is projected to reach 1.33 trillion CNY by 2029, with GPUs expected to account for 77.3% of this market [15][18] - The geopolitical risks have created historic opportunities for domestic companies, with predictions that local AI chip suppliers will capture 55% of the market by 2027 [15][18] Valuation and Risks - The market has priced in high expectations for Moole Thread's future growth, with a market capitalization that reflects a significant premium based on anticipated performance [19][20] - The company faces challenges such as profitability uncertainty, reliance on external funding, and a competitive landscape that has intensified with the entry of major players [23][24]

Core Viewpoint - The successful IPO of Moore Threads marks a significant milestone as it becomes the first domestic GPU company listed on the Sci-Tech Innovation Board, reflecting strong market interest and potential in the GPU sector [2][5]. Company Overview - Moore Threads was founded in 2020 and has quickly established itself in the GPU market, aiming to provide high-performance computing solutions for AI, digital twins, and scientific calculations [7][8]. - The company has developed a product matrix that includes multiple generations of GPUs, covering various applications such as AI computing, graphics rendering, and consumer entertainment [13]. IPO Details - Moore Threads was listed on December 5, with an opening price of 650 CNY per share, leading to a market capitalization of 305.5 billion CNY. The stock price later adjusted to 585 CNY, reflecting a 411.9% increase from the issue price [2][3]. - The IPO raised a net amount of 8 billion CNY, marking the largest scale for a Sci-Tech Innovation Board IPO in the past two years [3][4]. Financial Performance - The company reported revenues of 46.08 million CNY in 2022, with projections of 124 million CNY in 2023 and 438 million CNY in 2024, indicating a growth trend. However, net losses were recorded at 1.894 billion CNY in 2022, with expectations of narrowing losses in subsequent years [14][15]. - Moore Threads anticipates achieving profitability by 2027, supported by a commitment from its major shareholders to extend their lock-up period if profitability is not realized [14][16]. Market Position and Strategy - The company is positioned to compete directly with NVIDIA, leveraging its founder's experience in the GPU market to build a robust ecosystem around its products [6][8]. - Moore Threads aims to create a unified development environment through its self-developed MUSA architecture, which allows developers to work across its entire GPU product line [8]. Investment and Valuation - Prior to its IPO, Moore Threads attracted significant investment from various institutions, achieving a valuation exceeding 10 billion USD within 100 days of its establishment [9][10]. - The company has seen substantial returns for early investors, with some institutions reportedly making over 10 billion CNY from their investments [10][12].

Core Viewpoint - The article highlights the successful IPO of Moore Threads, marking it as the first domestic general-purpose GPU stock in China, with significant fundraising and market potential amidst the backdrop of increasing demand for AI chips and the exit of NVIDIA from the Chinese market [3][4][5]. Company Overview - Moore Threads, founded by Zhang Jianzhong, went public on the STAR Market with an issuance price of 114.28 CNY per share, raising 8 billion CNY, making it the largest IPO on the STAR Market in 2025 by fundraising amount [3][4]. - The company has a current market valuation of 280 billion CNY, with Zhang Jianzhong holding approximately 11% of the shares, translating to a personal wealth of 30.8 billion CNY [4][5]. Market Context - The exit of NVIDIA from the Chinese market due to U.S. export controls has created a significant gap in the AI chip market, with a projected shortfall of 16.8 billion USD in 2025 [5][39]. - Domestic AI chip companies, including Moore Threads, are facing challenges in meeting this demand due to supply chain limitations and the need for further technological advancements [5][39][42]. Financial Performance - Moore Threads has maintained high R&D investments, with projected losses of approximately 19 billion CNY, 17 billion CNY, and 16 billion CNY from 2022 to 2024, respectively [12][13]. - The company’s revenue has shown significant growth, with 2023 revenue reaching 785 million CNY, surpassing the total revenue of the previous three years [26]. Product Development - Moore Threads has developed a range of products, including general-purpose GPUs for AI computing and desktop GPUs, with a focus on rapid iteration and performance improvement [25][26]. - The latest product, the "Pinghu" S5000, has achieved a computing power of 32 TFlops, surpassing NVIDIA's A100 [25]. Competitive Landscape - The article discusses the emergence of four major domestic GPU companies, including Moore Threads, which are competing in the AI chip market, each with distinct focuses and product offerings [30][31]. - Despite the growth of domestic players, the market remains dominated by NVIDIA and AMD, with NVIDIA holding an 88% market share globally [48]. Investment and Support - Moore Threads has received substantial backing from government and venture capital, raising over 10 billion CNY prior to its IPO, with ongoing support from various institutional investors [12][19]. - The company has also benefited from government subsidies, totaling 585 million CNY over two and a half years [13][14]. Future Outlook - The article emphasizes that while the IPO is a significant milestone, it is merely the beginning of a long journey for Moore Threads and other domestic AI chip companies as they strive to compete with global giants [48].

Core Viewpoint - Moore Threads has officially listed on the Science and Technology Innovation Board, becoming the first domestic GPU stock, with a significant market response on its debut, achieving a market capitalization of approximately 305.5 billion yuan at opening [1][2]. Group 1: IPO Highlights - The IPO process for Moore Threads took only 88 days from acceptance to approval, setting multiple records [2]. - The company issued 29,382,386 new shares, raising a net amount of 8 billion yuan, marking the largest IPO scale on the Science and Technology Innovation Board in nearly two years [2]. - The initial offering price was 114.28 yuan per share, resulting in a 2024 diluted static price-to-sales ratio of 122.51 times, which is higher than the industry average but lower than Cambricon's 478.25 times [2]. Group 2: Company Background - Founded in 2020, Moore Threads is led by Zhang Jianzhong, a former NVIDIA executive who significantly increased NVIDIA's market share in China [3][4]. - The company aims to develop domestic GPUs to support AI infrastructure, with a goal of achieving profitability by 2027 [2][11][15]. - Moore Threads has built a strong team by recruiting core engineers from major tech companies like NVIDIA, Intel, AMD, and Microsoft [5]. Group 3: Product Development and Market Position - The company has developed a comprehensive product matrix, including four generations of GPUs, targeting various applications such as AI computing, high-performance computing, and gaming [12]. - Moore Threads' self-developed MUSA architecture allows developers to use a unified development environment across its GPU series, indicating a focus on creating a proprietary computing platform [6]. - The company is currently positioned primarily for domestic market replacement, competing directly with NVIDIA [7]. Group 4: Financial Performance - Moore Threads has shown a trend of increasing revenue, with reported revenues of 46.08 million yuan in 2022, 124 million yuan in 2023, and projected revenues of 438 million yuan in 2024 [12]. - Despite the revenue growth, the company has also reported significant net losses, which are gradually narrowing, with losses of 1.894 billion yuan in 2022 and 1.703 billion yuan in 2023 [12][14]. - The company has committed to substantial R&D investments, with expenditures of 1.116 billion yuan in 2022 and 1.334 billion yuan in 2023, representing a high ratio of R&D costs to revenue [11].

Core Viewpoint - The successful IPO of Moores Threads on the Shanghai Stock Exchange's Sci-Tech Innovation Board marks a significant milestone for the company, the board's reform, and the development of China's computing power industry [1] Group 1: Company Development - Moores Threads, known as the "first domestic GPU stock," aims to maintain an annual iteration speed for its full-function GPU chips [1] - The company achieved over 400 million yuan in revenue by 2024, with a compound annual growth rate of over 200% in the last three years [8] - The management expects the company to achieve profitability by 2027 [8] Group 2: Market Context - By 2029, intelligent computing power is projected to account for over 88% of China's total computing power [3] - The rapid IPO process of Moores Threads, taking only 158 days, reflects the supportive regulatory environment for high-tech industries in China [5] Group 3: Technological Innovation - Moores Threads has developed a unified system architecture called MUSA, which integrates GPU hardware and software, allowing compatibility with the international mainstream GPU ecosystem [11] - The company has defined its own "Moore's Law," with annual iterations of its GPU chips since its founding in 2020 [10] Group 4: Industry Insights - The development of the GPU industry is seen as a systemic project that requires advancements across the entire semiconductor supply chain, including chip design, manufacturing, and testing [12] - The company emphasizes the importance of policy support for the GPU chip industry and the semiconductor supply chain to enhance global competitiveness [13]

Core Insights - The core viewpoint of the article emphasizes the significance of the domestic GPU company, Moore Threads, as it successfully listed on the Shanghai Stock Exchange's Sci-Tech Innovation Board, marking a pivotal moment in China's computing power industry development [4][5][8]. Company Overview - Moore Threads, known as the "first domestic GPU stock," aims to leverage its listing to accelerate technological iterations and operational capabilities, with a goal to achieve profitability ahead of schedule [8][9]. - The company has introduced a series of full-function GPU chips since its establishment in 2020, with a projected revenue exceeding 4 billion yuan by 2024 and a compound annual growth rate of over 200% in the last three years [7][8]. Industry Context - The article highlights the increasing importance of computing power in the context of national strength, with GPU chips being foundational to this development [5][13]. - The Chinese government is actively supporting high-tech industries, particularly in artificial intelligence and semiconductor sectors, which is crucial for the growth of companies like Moore Threads [7][14]. Future Outlook - Moore Threads aims to become a leading GPU enterprise with international competitiveness, focusing on creating advanced computing acceleration platforms for AI and digital twin applications [12][13]. - The company plans to invest in three major projects post-IPO, including AI training and inference chips, next-generation graphics chips, and AI SoC chips, to support a wide range of computing needs [12].