Core Viewpoint - The article highlights the rapid growth and strategic moves of Zhiyu Technology, which has recently seen its stock price surge over 138.68% in just five trading days, marking it as a significant player in the AI sector and the first "large model" stock in Hong Kong [1]. Group 1: IPO Developments - Zhiyu Technology has updated its A-share IPO guidance report, adding Guotai Junan Securities as a new advisor alongside CICC, indicating an upgrade in its IPO advisory team [3][4]. - The company had initially aimed for an A-share listing earlier but shifted its strategy to list on the Hong Kong Stock Exchange first, which it achieved in January 2026 [6]. - The inclusion of Guotai Junan Securities is expected to accelerate Zhiyu's A-share IPO process, leveraging the firm's strong track record in the tech sector [6][7]. Group 2: Competitive Landscape - The competition among leading securities firms for IPO services in the AI sector is intensifying, with firms like CICC and CITIC Securities holding significant positions in both Hong Kong and A-share markets [10]. - Guotai Junan Securities has been involved in 19 IPO projects on the Sci-Tech Innovation Board over the past three years, making it the top firm in terms of project count, compared to CICC's 12 projects [6][7]. - The article notes that the collaboration between CICC and Guotai Junan Securities is not new, as they have previously worked together on significant IPOs in the "hard tech" sector [8][9]. Group 3: Market Trends and Challenges - The current market environment is favorable for AI companies seeking IPOs, but the complexity of their technologies and business models means that they face rigorous scrutiny during the application process [11]. - The article emphasizes that while the policy environment has opened doors for unprofitable "hard tech" companies, the requirements for high-quality submissions remain stringent [11].

Core Viewpoint - The stock price of Zhiyuan Technology, known as the "first stock of large models" in Hong Kong, surged over 138.68% within five trading days, indicating strong market interest and momentum for AI-related companies [2][3] Group 1: IPO Developments - Zhiyuan Technology has updated its A-share IPO guidance report, adding Guotai Junan Securities as a new advisor alongside CICC, enhancing its advisory team [3][5] - The company previously withdrew its IPO guidance submitted in April 2025 and is now aiming for a listing on the Shanghai Stock Exchange's Sci-Tech Innovation Board [5][6] - The rapid re-initiation of A-share IPO guidance reflects Zhiyuan's commitment to establishing a dual capital platform ("A+H") [7] Group 2: Competitive Landscape - The competition among leading securities firms for IPO services in the AI sector is intensifying, with a focus on both speed and quality of submissions [3][12] - Guotai Junan Securities has a strong track record, having participated in 19 IPO projects on the Sci-Tech Innovation Board over the past three years, making it the top firm in terms of project count [8] - In contrast, CICC has been involved in 12 projects during the same period, highlighting the competitive dynamics between these top-tier firms [9] Group 3: Market Trends - The current environment allows unprofitable "hard tech" companies, including AI firms, to pursue IPOs, but the complexity of their technologies and business models leads to more rigorous scrutiny during the application process [14] - The collaboration between CICC and Guotai Junan Securities in past projects, such as the successful IPO of SMIC, showcases their established partnership in the hard tech sector [11][10]

上海营券交易所 SHANGHAI STOCK EXCHANGE TE H दुरुं FIT 2 9 2025上交所市场运行回顾 米 ma ■ 2.444 ♀ 上证指数 涨幅 o l8.4% 科创综指 � 涨幅 46.3% © 聚焦服务新质雪音力 科创成长层新上市禾元生物、摩尔线程、沐曦股份等 6家企业 科创成长层已汇聚39家企业(截至2026年2月13日) 同比增长 科创板新增股权融资 1073亿元 2.8倍 并购市场活力进发 新增披露资产重组 其中: 重大资产重组 9|4家次 95家次 同比增长 3 830% 大型吸收合并、多元支付等创新示范性项目相继涌现 沪市围绕新质生产力的并购重组占比 近7成 债券市场提质扩面 H 债券托管量 (全球最大的交易所债券市场之一) 19.5万亿元 沪市债券新增发行 同比增长 (6) 7.9万亿元 11% 同比增长 其中: 产业债新增发行 36% 2.1万亿元 科创债新增发行 同比增长 超9000亿元 66% O REITs融资规模 345亿元 持有型不动产ABS产品发行规模 C 500亿元 O 经营业绩稳中有升 n | 前三季度沪市公司 (●) 营收37.6万亿元 其中: ...

Core Viewpoint - Shanghai Suiruan Technology Co., Ltd. is focusing on cloud AI chip design and has recently changed its IPO review status to "inquired" on the Sci-Tech Innovation Board, aiming to raise approximately 6 billion yuan for the development and industrialization of its AI chip series products [1][6]. Group 1: Company Overview - Suiruan Technology was established in 2018 and is recognized as one of the leading domestic cloud AI chip companies, alongside other firms like Moore Threads and Biran Technology, collectively referred to as the "Four Little Dragons of Domestic GPUs" [1][2]. - The company has developed four generations of cloud AI chips and has built a complete product system including AI chips, AI accelerator cards, intelligent computing systems, and AI computing software platforms [2][3]. Group 2: Market Position and Competition - The domestic cloud AI chip industry is in its early stages, with international players like NVIDIA holding a significant market share, while local companies are gradually breaking through technological barriers [1][6]. - Suiruan Technology's latest fourth-generation cloud AI chip supports FP8 low-precision data and has performance metrics that can compete with international products, indicating that its core technology has reached a leading level domestically [3]. Group 3: Financial Performance - The company's revenue has shown rapid growth, with a compound annual growth rate of 183.15% from 2022 to 2024, although it has not yet achieved profitability [6]. - Revenue figures for the years 2022, 2023, 2024, and the first nine months of 2025 are 90.1 million yuan, 300 million yuan, 720 million yuan, and 540 million yuan respectively, with net losses of 1.15 billion yuan, 1.567 billion yuan, 1.503 billion yuan, and 912 million yuan [5][6]. Group 4: Investment and Future Plans - The IPO proceeds will primarily be invested in the research and industrialization of the fifth and sixth generation AI chip series, as well as advanced AI hardware and software projects, with total planned investments of 6 billion yuan [6]. - The company has established a strong partnership with major internet firms, which has led to widespread commercial applications of its products in various AI scenarios [5].

Group 1 - The core idea of the article revolves around the intense competition among technology companies for high-end talent, highlighted by extravagant year-end bonuses including houses, cars, and gold [4][6][10] - Liu Jingkang, CEO of Yingshi, emphasized that material incentives are crucial for attracting top talent, alongside a sense of achievement and honor [3][4] - The article notes that many tech companies have successfully raised funds and are using idle capital to invest in financial products from banks and brokerages, with some companies utilizing up to 99% of their financing for such investments [3][13] Group 2 - Yingshi's year-end bonuses included five apartments, five cars, a Porsche, and gold, showcasing the company's commitment to attracting talent [1][7] - Other companies like Tuo Zhu Technology and SK Hynix are also offering competitive year-end bonuses, with SK Hynix providing an average of 640,000 yuan per employee [7][12] - Tencent's year-end bonuses are linked to performance, with some employees reportedly receiving bonuses equivalent to 30 months of salary, indicating a strong financial position [8][9] Group 3 - The article discusses the broader trend of tech companies engaging in a "talent war," where financial resources play a critical role in attracting skilled professionals [10][12] - Companies like Meta are offering substantial compensation packages to attract AI researchers, with reports of salaries reaching up to $3 million over four years [12][13] - The demand for talent in the GPU and AI sectors is high, with a significant talent gap leading to increased salary premiums for experienced engineers [13][14]

作为我国经济中心、科技创新高地与对外开放前沿，近年来，上海聚力打造全国领先的智能算力高地， 推动智算资源加速集聚、产业生态持续迭代提质。上海市算力网络协会数据显示，上海智算规模在全国 占比约达8%，已成为引领全国智算产业高质量发展、带动生态协同升级的重要增长极。 中国证券报记者调研发现，从智算规模的跨越式突破到优质企业的集群式崛起，从技术架构的根源性创 新到产业发展共识的深度凝聚，从生态要素的零散分布到全链条协同发力，上海正稳步突破传统规模扩 张、资本驱动的发展瓶颈，引领智算产业实现从"规模增速"向"质量引领、效率优先、可持续发展"的高 阶转型，同步推动智算生态开放共享、协同共生，以标杆性的"上海实践"赋能国家智算产业能级跃升。 天数智芯AI与加速计算技术负责人单天逸把未来的高质量算力定义为"高效率、可预期、可持续"。"所 谓高效率，是指能为客户创造最优的TCO(总体拥有成本)，实实在在帮客户节省使用成本。而可预期是 能够通过精准的仿真模拟，让客户在拿到芯片、部署算力之前，就能清晰预判最终的性能表现，做到所 见即所得。至于可持续，则是从现在主流的算法再到未来还未诞生的全新算法，算力始终能无缝适配， 今天能用， ...

Core Insights - Shanghai is positioning itself as a leading hub for intelligent computing power, contributing approximately 8% to the national scale of intelligent computing, and driving high-quality development in the industry [1][2] - The average annual compound growth rate of China's intelligent computing scale is projected to reach 46.2% from 2023 to 2028 [1] Industry Development - Shanghai's intelligent computing scale has surpassed 120,000 PetaFLOPS, with nearly 20,000 PetaFLOPS managed by the municipal scheduling platform [2] - The industry is shifting focus from capital-driven growth to quality, efficiency, and sustainability, emphasizing the need for deep value cultivation and innovation [2][3] - The market is transitioning from measuring computing power in FLOPS to evaluating application effectiveness, marking a new phase of high-quality development [3] Innovation and Technology - Innovation is identified as the primary driver for high-quality development in the intelligent computing industry, with local companies like MuXi, BiRan Technology, and TianShu ZhiXin making significant breakthroughs [4][6] - MuXi has launched the X series GPU aimed at enhancing performance in scientific computing, supporting various complex simulations [5][6] - TianShu ZhiXin is focusing on creating a high-quality computing architecture with a three-year roadmap, aiming to deliver high-performance and cost-effective solutions [6] Ecosystem Collaboration - The development of the intelligent computing ecosystem in Shanghai is characterized by collaborative efforts among industry players, creating a network of shared value [7][8] - Companies like UCloud are playing a pivotal role in building a collaborative infrastructure that integrates high-performance computing with local industry needs [7][8] - The "Shanghai Intelligent Computing Industry High-Quality Development Initiative" was launched to promote the construction of new intelligent computing centers and a high-level cloud service system [9][10]

Core Viewpoint - The article highlights the rapid development and competitive landscape of domestic AI chip companies in China, particularly focusing on the six emerging players in Shanghai, which are poised for significant growth and innovation in the AI sector [2][10]. Group 1: Company Developments - Hanbo Semiconductor has completed its guidance for an IPO on the Sci-Tech Innovation Board and is progressing steadily towards its public offering [2]. - Suiruan Technology's IPO application has been accepted, indicating a growing interest in AI chip companies [2]. - The six domestic AI chip companies, including Hanbo Semiconductor, are collectively referred to as the "Six Little Dragons," with a strong emphasis on technological advancements and product development [2][4]. Group 2: Market Position and Strategy - The "Six Little Dragons" are primarily cloud-based AI chip manufacturers, with five of them focusing on GPGPU architecture to compete with global giants like NVIDIA and AMD [2][4]. - Hanbo Semiconductor specializes in cloud AI inference and has achieved commercial success in both general AI and rendering applications, positioning itself as a leader in the domestic market [3][4]. - The companies are actively updating their technology and product lines, with significant launches such as the new GPU brand by Muxi Co., which aims to enhance scientific research capabilities [3][4]. Group 3: Financial Performance and Projections - Muxi Co. anticipates a revenue of 1.6 billion to 1.7 billion yuan in 2025, reflecting a year-on-year growth of 115.32% to 128.78% [9]. - Moer Thread expects a revenue of 1.45 billion to 1.52 billion yuan in 2025, with a projected year-on-year increase of 230.70% to 246.67% [9]. - The domestic general GPU market share is expected to rise from 8.3% in 2022 to over 50% by 2029, indicating a structural transformation in the market [10]. Group 4: Government and Investment Support - Shanghai has been proactive in nurturing the AI chip industry through financial support and favorable policies, positioning itself as a leader in the integrated circuit sector [5][7]. - The Shanghai Sci-Tech Innovation Fund has been involved in early-stage investments in companies like Muxi, demonstrating strong government backing for local tech firms [6]. - Hanbo Semiconductor has received significant investments from local state-owned enterprises, benefiting from various policy supports aimed at fostering high-quality development in the tech sector [7].

Core Viewpoint - Recently, two companies, Muxi Integrated Circuit (Hangzhou) Co., Ltd. and Muxi Qizhi Technology (Taizhou) Co., Ltd., have announced their dissolution due to a resolution to disband, with the announcement period from February 5 to March 21 [1] Company Information - Muxi Integrated Circuit (Hangzhou) Co., Ltd. was established in December 2020 with a registered capital of 25 million RMB, while Muxi Qizhi Technology (Taizhou) Co., Ltd. was founded in August 2024 with a registered capital of 50 million RMB [1] - Both companies are wholly owned by Muxi Co., Ltd. (688802) and have business scopes that include integrated circuit design, software development, and software sales [1]

Core Insights - The article emphasizes the significant challenge of ecological barriers faced by domestic GPU companies in China, which goes beyond hardware performance and capital market achievements [1] - The current global GPU market is dominated by Nvidia and AMD, whose competitive advantage has shifted from hardware specifications to software and ecosystem development [1] - The article highlights the importance of building a robust ecosystem that requires deep understanding of developer needs, continuous improvement of the toolchain experience, and long-term community cultivation [2] Group 1: Ecological Barriers - Domestic GPU manufacturers are adopting a "compatible ecosystem" strategy to quickly enter the market, but this may lead to a long-term dependency on existing ecosystems, hindering the development of unique competitive advantages [2] - The construction of an ecosystem is a slow process that tests patience, capital, and foresight, requiring solid technical foundations and ongoing support from investors [2] Group 2: Industry Constraints and Capital Challenges - Domestic GPU companies face multiple constraints from the industry foundation, including limitations in advanced manufacturing processes, core IP self-sufficiency, and challenges in advanced packaging technologies [3] - The high investment nature of the GPU industry makes it a "capital endurance battle," with companies like Wallen Technology investing 85% of their IPO proceeds into R&D [3] - The financial performance of domestic GPU companies often shows strong revenue growth but delayed profitability, reflecting the industry's inherent characteristics [3] Group 3: Market Survival and Competition - The current growth of domestic GPUs is largely driven by policy support and specific replacement demands, but true commercial success requires navigating a fully market-driven and competitive landscape [4] - The ultimate test for domestic GPUs lies in their ability to support large-scale AI models, run top-tier games, and provide reliable services in high-precision industrial simulations and life sciences [4] Group 4: Investment Value Reevaluation - Investors are advised to maintain a rational perspective on the GPU sector, focusing on actual penetration rates, order visibility, and the quality of revenue growth [7] - Key mid-term indicators include breakthroughs in core architecture, the robustness of the software ecosystem, and clarity in profitability pathways [7] - Long-term success will depend on the ability to create a globally influential development platform and transition from a technology follower to a rule-maker [7] Group 5: Strategic Importance of GPU Development - The evolution of GPUs represents a strategic contest for control over foundational infrastructure and standards in the intelligent era, with significant implications for China's technological competitiveness [9] - The collective efforts of domestic GPU companies are seen as a crucial expedition that aligns with national expectations and industrial missions [9] - The journey towards establishing a self-sufficient GPU industry is expected to be challenging yet rewarding, with the potential to reshape the global computing landscape [10]