在国产GPU四小龙接连登陆资本市场的同时，国产算力芯片也进入大规模产业化落地的关键时期。现 场，施淑珏还谈到沐曦在商业化方面的探索，以及新一年的发展规划情况。 商业化核心门槛在于从"可用"到"好用"的跨越 施淑珏表示，沐曦已构建了从硬件架构到软件栈的完整自研技术体系。凭借全建制的研发团队，沐曦得 以推出覆盖人工智能训练和推理、通用计算、图形渲染的全栈GPU产品矩阵。目前，沐曦已具备从芯 片、到板卡、服务器、超节点直至部署大规模算力集群的全链条能力。 近年来，国产GPU的落地应用情况成为业界重点关注的话题。施淑珏指出，国产GPU商业化的核心门槛 在于从"可用"到"好用"的跨越。在她看来，衡量一款GPU是否好用，除了单芯片算力，也需要多维度的 能力，沐曦产品在通用性、易用性、稳定性和规模可扩展性四个方面均有亮眼的表现。 "我们更加注重真正交付给客户在AI应用场景实战中的产品性能。"她说。 近年来，在人工智能浪潮席卷全球、算力需求呈指数级增长的背景下，国产GPU成为资本市场瞩目的焦 点。2025年12月，摩尔线程、沐曦股份的接连科创板上市，再度将市场对国产GPU的追逐推向高潮。 近日，胡润研究院在北京亦庄发布的《2 ...

21世纪经济报道记者申俊涵北京报道 近年来，在人工智能浪潮席卷全球、算力需求呈指数级增长的背景下，国产GPU成为资本市场瞩目的焦点。2025年12月，摩尔 线程、沐曦股份的接连科创板上市，再度将市场对国产GPU的追逐推向高潮。 近日，胡润研究院在北京亦庄发布的《2025胡润中国人工智能企业50强》显示，寒武纪、摩尔线程、沐曦股份成为排名前三的 人工智能企业。据悉，该榜单按照企业价值进行排名，上市公司市值按照2026年1月9日的收盘价计算，非上市公司估值参考同 行业上市公司或者根据最新一轮融资情况进行估算。 "我们更加注重真正交付给客户在AI应用场景实战中的产品性能。"她说。 据悉，沐曦是国内少数实现千卡集群大规模商业化应用的GPU供应商，正研发和推动万卡集群的落地。施淑珏认为，这考验的 是中国企业最擅长的"工程能力"。事实上，中国企业在芯片设计方面，已经能够跟国际企业进行同等对话。 其中，寒武纪凭借 6300亿元的价值，位居榜单首位。2025年上半年，寒武纪实现营收28.8亿元，同比大幅增长43倍；摩尔线程 以3100亿元的价值排名第二，2025年公司以创科创板IPO最快过会纪录——88天。2025年前三季度 ...

在国产大算力芯片迈入大规模产业化落地的关键周期，头部玩家的座次正在重排不断实现自我突破。 近日，胡润研究院在北京亦庄发布《2025胡润中国人工智能企业50强》。在这份衡量中国AI产业成色 的重磅榜单中，沐曦集成电路（下称"沐曦"）以2500亿元人民币的企业价值位列第三，并被官方定义 为"中国首批实现全流程国产化的高端GPU（图形处理器）企业"。 软件生态"即插即用"：MACA开源一年调用超1300万次 硬件是底座，软件则是护城河。为了打破算力领域的软件垄断，沐曦自研了原生兼容主流生态的软件栈 MXMACA。 施淑珏透露，MXMACA已于2025年2月正式开源，在不到一年的时间里，该平台已积累了超过15万用 户，API调用次数突破1300万次。通过与高校合作，沐曦的软件生态已覆盖超过10万名大学生开发者。 "原生兼容意味着AI应用可以几乎零成本地迁移到沐曦平台上，实现'即插即用'。"施淑珏分享道。 "1+6+X"策略：助力AI赋能千行百业金融领域已现"三期复购" 在应用落地层面，沐曦提出了"1+6+X"的行业布局方案。其中，"6"涵盖了金融、医疗、能源、教科 研、交通及大文娱六大垂直行业；"X"则指向具身智能、 ...

Summary of Key Points from the Conference Call Industry Overview - The domestic AI chip market is expected to face a supply shortage in 2026, with the gap between domestic and foreign chip supply narrowing due to increased domestic production capacity and gradual opening of supply from Europe and the US [1][2] - The Internet Data Center (IDC) sector is showing signs of recovery, with large-scale bidding activities resuming, indicating optimism about future demand [1][2] Core Insights and Arguments - Major domestic AI models are expected to close the gap with overseas counterparts, with a projected supply-demand imbalance not only in chips but also in AI models themselves [2] - The introduction of new domestic computing products, such as Huawei's Ascend 950PR, is anticipated in Q1 2026, which may drive more orders from internet companies [1][2] - Two significant AI model companies, Zhiyu and MinMax, are expected to go public around the Spring Festival, which could enhance market confidence in domestic AI enterprises [2] Growth Projections - Domestic computing power is projected to grow by over 50% year-on-year in 2026, driven by increased capital expenditure from domestic internet companies, improved chip performance, and supportive policies [1][3] - The AIDC (AI Data Center) industry chain is experiencing significant changes, with a surge in bidding activity in the first half of 2025, although supply constraints limited growth in the second half [1][7] New Product Launches - The 690 chip is expected to be sampled in early 2026, with its performance and market feedback being crucial for its success [1][5] - Other notable products include Haiguang's DCO 4 and the upcoming C600 and C700 from Muxi, which are expected to be released in 2026 [5] Market Influencing Factors - Increased capital expenditure from domestic internet companies and the performance improvements of chips are key drivers of market growth [3][4] - Policy changes, such as the approval process for high-performance computing cards, will significantly impact market dynamics [4] Beneficiaries of AIDC Industry Changes - Companies with significant internet exposure, such as Chai Fa, are expected to benefit from potential price increases [8] - Other companies to watch include server manufacturers like Inspur and those involved in ITC bidding, such as Runze and Data Port [8] Impact of New Market Entrants - The entry of new players like Enka is not expected to negatively impact domestic cards due to the ongoing supply-demand tightness in 2026 [9] - Focus should be on companies that can continue to raise prices or those that follow volume growth logic, including supply chain players like Weichai and Ruijie [9]

Summary of Conference Call on Huawei's Super Node and Domestic AI Chip Market Industry Overview - The conference call discusses the domestic AI chip market, focusing on Huawei's 384 Super Node and its competitive landscape against international players like NVIDIA [1][2][8]. Key Points and Arguments - **Huawei 384 Super Node**: - Utilizes high-speed interconnect and optical module technology, enhancing cost-effectiveness and deployment certainty in inference scenarios, serving as a domestic alternative to NVIDIA's Nvlink bottleneck [1][2]. - Supports SP8, providing superior performance in inference tasks compared to many domestic cards lacking this feature [2]. - **Domestic AI Chip Companies**: - Companies like Moer, Muxi, and Suiyuan are narrowing the gap with international leaders through technological advancements, with expectations to increase market share by 2026 [1][4]. - New products from these companies, such as Moer S5000 and Muxi C600, are expected to meet the demands of major internet companies, enhancing their competitive position [4]. - **Market Demand and Supply**: - The demand for domestic computing cards is not limited to internet companies but extends to telecom operators, financial institutions, and state-owned enterprises, indicating a supply-demand imbalance [3][16]. - The market is projected to focus on 2026 valuations, with optimism surrounding the growth of domestic graphics cards [3][13]. - **High-Density Super Point Solutions**: - Companies like Haiguang are developing high-density, high-performance super point solutions, such as a silent liquid cooling solution supporting up to 660 cards, to improve overall computing efficiency [5][6]. - **GPGPU Architecture Development**: - Domestic GPGPU architecture is evolving, with companies like Huawei Ascend transitioning to GPGPU to reduce interconnect development risks [7]. - The integration of DSA functions into GPUs by manufacturers like Moer indicates a trend towards technological convergence [7]. - **Future Market Dynamics**: - The competitive landscape is expected to become more complex by 2026, with new product launches and testing processes for internet companies [10][12]. - Despite the anticipated changes, the overall market structure is expected to remain stable, with Huawei leading, followed by Cambricon and Haiguang [10]. Additional Important Insights - **Inference Demand Growth**: - The increasing demand for inference capabilities, especially in the context of the multi-modal era, is driving the need for domestic computing cards [9][12]. - **AI Application Revenue**: - Significant growth in AI application revenues is anticipated, with estimates for companies like Keling reaching $200 million in 2025 [10]. - **Comparison with U.S. Market**: - The U.S. market is dominated by a few major clients, while China's demand is diversified across numerous long-tail customers, indicating a robust growth potential for domestic computing cards [16]. This summary encapsulates the critical insights from the conference call, highlighting the competitive landscape, technological advancements, and market dynamics within the domestic AI chip industry.