2025年，注定是不凡的一年。回望这一年，科技领域以惊人的速度重塑着我们的生活，以人工智能、商 业航天等为代表的硬核科技深刻重塑着经济社会发展的底层逻辑与未来图景，"未来"正在加速度撞进现 实。 技术浪潮不仅催生了新的生产力，更成为驱动资本市场结构性行情的核心引擎。从底层算力设施的迭代 到前沿应用的普及，一系列年度科技热词清晰地勾勒出产业跃迁的路径，也为洞察未来投资方向提供了 关键坐标。 在AI算力需求的推动下，作为电子设备"骨架"的高端印刷电路板(PCB)产业实现技术跃迁。产品向高层 数、高密度、高可靠性方向加速演进，单板价值量与技术壁垒显著提升，这个传统产业正在成为支撑电 子信息产业创新发展的基石。正因如此，2025年PCB产业的强势崛起与技术蜕变，也标志着中国在全球 高端制造版图中，正从一个规模领先的"制造中心"，向定义技术标准的"创新源"与"赋能极"稳步跨越。 年度热词透视：技术突破与资本共振轨迹 这一年，从AI底层算力到太空前沿探索，每个"热词"都对应着一条崛起的产业链，而不少也映射到了市 场行情中。 热词一：AI智能体——AI的"觉醒时刻" 2025年初，以DeepSeek为代表的大模型作出战略性 ...

她在致辞中梳理了华为过去一年的主要工作成果，包括5G-A建设、鸿蒙生态体验、乾崑智驾，计算产 业中的鲲鹏、openEuler、昇腾和超节点，以及行业AI落地和绿色能源等。 孟晚舟表示，面向未来，智能化变革浪潮奔涌，这是我们面临的长期战略机遇。 她总结了华为2026年的"主战场"： 12月30日，华为轮值董事长孟晚舟发表2026年新年致辞。 始终从消费者体验出发，繁荣鸿蒙生态，畅享AI体验，为终端消费者带来更多惊喜； 鸿蒙智行和乾崑智驾助力车企规模上量，打造安全舒适的驾乘体验； 重构AI数据中心，让每瓦特产出更多Tokens； 液冷超快充，让有路的地方就有高质量的充电体验； …… "过去的成功并非未来的航标，我们深信，在充满不确定性的智能化远航中，最终的胜利，属于那些敢 于挺身而出、勇于自我批判的奋斗者。战略聚焦，决胜价值战场，以质取胜，做强组织能力，2026年的 我们，必将全力以赴！"孟晚舟写道。 强化行业垂直作战，发挥"大杂烩"优势，深耕行业，使能千行百业智能化转型； 构建开源开放的鲲鹏昇腾生态，使能伙伴开发满足各行各业需求的产品，推动集群与超节点 技术普惠，构筑坚实的AI算力底座； "水战略"激发管道流量 ...

（编辑 丛可心 王雪儿） 证券日报网讯 12月24日，华勤技术在互动平台回答投资者提问时表示，华勤技术始终高度重视研发能 力建设与技术创新，在研发端保持持续而稳健的投入，2025年前三季度研发费用合计46.20亿元，同比 增长23.7%，预计全年研发投入超过60亿元人民币。目前公司研发技术人员将近2万名，研发投入方向 主要是围绕3+N+3的产品布局以及满足业务增长的需要；同时，公司也会投入一些前瞻性的研发预研， 如Xlab在声学、光学、热学、射频等领域的研究；针对新技术和新产品方向如AI端侧、超节点、汽车电 子等方向都持续增加研发资源；机器人也是公司一个明确的投入方向，公司会继续保持稳健的投入；研 发是公司的核心竞争力，公司将持续加大研发投入，夯实技术护城河，实现可持续高质量发展。 ...

Group 1 - The first Artificial Intelligence Innovation Conference (HAIC2025) was held in Kunshan, Jiangsu, gathering over 2,500 representatives from various industries, experts, and leaders to showcase China's AI computing open architecture and ecological prosperity [1] - AI has transitioned from an auxiliary tool to a core production factor, with significant advancements in scientific models and embodied intelligence, leading to unprecedented demands for computing power [2][3] - Challenges such as insufficient high-end computing supply, poor hardware-software compatibility, inconsistent technical standards, and high application costs are significant barriers for small and medium enterprises [2][3] Group 2 - The conference emphasized the need for an open, efficient, secure, and stable intelligent computing infrastructure, advocating for a collaborative and win-win industrial pattern [2][3] - Notable innovations presented included the scaleX supercluster, designed for trillion-parameter models, which consists of 16 scaleX640 supernodes interconnected by a high-speed network, marking a key breakthrough in building world-class large-scale intelligent computing infrastructure [5] - Strategic collaborations were established among companies like SenseTime, Inspur, and others to optimize AI computing hardware and software systems, focusing on innovations in world models and embodied intelligence applications [5]

利空落地。 12月10日，中科曙光开盘瞬间跌停，海光信息一度跌超5%。消息面上，就在前一晚，海光信息、中科曙光宣布终止千亿元换股合并计划。 这一千亿级别的重大资产重组的终止，影响究竟有多大？券商中国记者采访发现，券商分析师普遍认为，重组的终止对于中科曙光而言会有一定冲击，而对于海光 信息而言更多的是利空落地，影响较小。更有券商提出，"站在未来看，今天或许是最佳买点！" "终止"影响有多大？ 12月9日晚间，海光信息、中科曙光同时发布公告，宣布终止重大资产重组，主要原因是交易规模大、涉及相关方较多，且市场环境较筹划之初发生较大变化。两 家公司均表示，本次交易终止不会对公司的生产经营和财务状况造成重大不利影响，不存在损害公司及中小股东利益的情形，并承诺至少1个月内不再筹划重大资 产重组事项。 刘熹还表示，对于交易中科曙光换股海光信息20%溢价的资金而言，截至12月9日收盘，按1:0.5525换算，中科曙光换股海光信息的价差为21%（套利空间），而在 重组预案出台前（6月9日前）该价差为21.5%，价差已从最低的7.62%（6月11日）回到了重组事件前。这意味着，市场已提前预期了重组终止风险。 一位券商电子行业首席 ...

Group 1: Industry Overview - The computer industry index increased by 18.54% from January 1, 2025, to November 28, 2025, outperforming the CSI 300 index by 3.50 percentage points [3][17] - In the first three quarters of 2025, the computer industry generated revenue of 935.84 billion yuan, a year-on-year increase of 9.14%, while net profit attributable to shareholders was 23.14 billion yuan, up 31.95% year-on-year [3][24] - The third quarter of 2025 saw revenue of 324.49 billion yuan, a 4.75% year-on-year growth, with net profit increasing by 23.30% to 10.26 billion yuan [3][24] Group 2: AI Computing Power - AI computing power demand is driving growth in the global data center market, which is expected to reach 652.01 billion dollars by 2030, growing at a CAGR of 8.53% from 2018 [35] - China's data center market is projected to grow from 68.01 billion yuan in 2018 to 261.33 billion yuan by 2024, with a CAGR of 25.15% [35] - The introduction of supernode technology is expected to break through computing power bottlenecks, enhancing efficiency and scalability in AI applications [62][66] Group 3: AI Models - Global large model iterations are accelerating, with domestic manufacturers focusing on open-source models to narrow the gap with overseas closed-source models [69] - Domestic open-source models are gaining traction, with eight out of the top ten in the global open-source model intelligence index being Chinese [73][76] - The performance gap between domestic open-source models and international top-tier closed-source models is narrowing, with domestic models offering better cost-performance ratios [76] Group 4: AI Applications - AI applications are transitioning towards intelligent agents, with major internet companies accelerating the commercialization of C-end applications [82] - Alibaba's Qwen App, based on open-source models, aims to become a comprehensive AI life portal, achieving over 10 million downloads shortly after launch [83] - The development of AI agents is expected to enhance user engagement and create a complete AI business ecosystem [83] Group 5: Investment Strategy - The report suggests focusing on opportunities in AI applications and self-control developments for 2026, with a positive outlook on the AI computing power sector [84][85] - Continuous growth in capital expenditure from major cloud computing firms supports the high prosperity of the AI computing power industry [84] - The competition in the global large model field is intensifying, which is likely to drive technological innovation and accelerate application deployment [84]

Group 1 - Huawei is set to release a breakthrough technology in the AI field on November 21, which aims to improve the efficiency of computing resource utilization from the industry average of 30%-40% to 70% [1] - The new technology will enable unified resource management and utilization of computing power from various sources, including GPUs and NPUs, enhancing support for AI training and inference [1] - Huawei's AI chip roadmap includes three series of products: 950PR/950DT, 960, and 970, with the 950PR expected to launch in Q1 next year, featuring enhanced interconnect bandwidth and computing power [1] Group 2 - Analysts from Dongfang Securities believe that the future of computing power competition will shift from GPUs to supernodes, which will become the mainstream form of computing power [2] - The emergence of large-scale supernodes, such as those showcased by Nvidia and Huawei, is expected to significantly enhance the training efficiency of domestic large models in China [2] - The domestic computing power market is anticipated to experience a boom in both large model training and AI capital expenditure next year [2] Group 3 - Tuowei Information is a strategic partner of Huawei in the "Cloud + Kunpeng/Ai + Industry Large Model + Open Source Harmony" field, developing innovative solutions based on the Harmony operating system [3] - Huafeng Technology, as a high-speed line module supplier, has established deep partnerships with leading domestic AI server manufacturers, including Huawei, ZTE, and Alibaba [3]

Core Viewpoint - The article discusses the emergence and significance of the "SuperNode" concept in the AI computing market, highlighting the competitive landscape among domestic manufacturers aiming to match or surpass Nvidia's offerings [1][11]. Group 1: SuperNode Concept - The term "SuperNode" refers to high-performance computing systems that integrate multiple AI training chips within a single cabinet, enabling efficient parallel computing [5][7]. - Domestic manufacturers have rapidly adopted the SuperNode concept, with various companies showcasing their solutions at industry events, indicating a collective push towards advanced AI computing capabilities [2][4]. Group 2: Performance Metrics - Companies are emphasizing the performance metrics of their SuperNode products, with Huawei's 384 SuperNode reportedly offering 1.67 times the computing power of similar Nvidia devices [3][12]. - The scale of integration, indicated by numbers like "384" or "640," reflects the number of AI training chips within a single system, serving as a key performance indicator for manufacturers [7][8]. Group 3: Challenges and Solutions - The industry faces a "communication wall" where a significant portion of computing time is spent waiting for data transfer, necessitating the development of SuperNodes to enhance communication efficiency [6][9]. - The transition from traditional computing methods to SuperNode architectures is driven by the need for higher performance in training large AI models, with manufacturers exploring both Scale-Up and Scale-Out strategies [7][8]. Group 4: Competitive Landscape - Domestic firms are positioning their SuperNode products against Nvidia's offerings, with Huawei's Atlas950 expected to outperform Nvidia's NVL144 in several key metrics [11][12]. - The competition is not only about performance but also about innovative engineering solutions to manage power consumption and heat dissipation in densely packed systems [13][15]. Group 5: Market Demand - The primary demand for AI computing resources is expected to come from large internet companies and state-led cloud services, which are likely to drive the market in the next few years [20][21]. - There are concerns about the sustainability of this demand, as companies may face challenges in justifying high capital expenditures for advanced computing resources [21][22]. Group 6: Future Outlook - The article suggests that while hardware challenges exist, the real test for domestic manufacturers will be in developing robust software ecosystems to support their SuperNode offerings [19][22]. - There is optimism about the potential for AI applications in sectors like robotics and advanced manufacturing, which could drive sustained demand for high-performance computing solutions [22].

Core Insights - The AI computing power market is increasingly focused on "SuperNode" technology, with multiple companies showcasing their solutions at various conferences throughout 2023 [2][3] - The emergence of SuperNodes is driven by the need to overcome bottlenecks in training large AI models, particularly the "communication wall" that arises during parallel computing [4][9] - Domestic companies are adopting SuperNode technology as a practical solution to enhance overall computing power, compensating for limitations in single-chip performance [10][12] Group 1: SuperNode Technology - SuperNode refers to a high-density computing solution that integrates multiple AI chips within a single cabinet, allowing them to function as a unified system [6][7] - The design of SuperNodes involves two main approaches: Scale-Up, which increases resources within a single cabinet, and Scale-Out, which connects multiple cabinets [5][8] - The numbers associated with SuperNodes (e.g., "384", "640") indicate the number of AI training chips integrated within a single system, serving as a key metric for performance and density [7][8] Group 2: Industry Competition - Companies like Huawei and Inspur are positioning their SuperNode products as superior to NVIDIA's offerings, with Huawei claiming its Atlas 950 will outperform NVIDIA's NVL144 in multiple performance metrics [10][11] - The competitive landscape is marked by aggressive parameter comparisons, with domestic firms striving to achieve higher integration density within their SuperNode solutions [12][14] - The engineering challenges of integrating numerous high-power chips into a single cabinet necessitate advanced cooling and power supply technologies [12][14] Group 3: Market Demand and Challenges - The primary demand for AI computing power is expected to come from large internet companies and state-led cloud services, which have the infrastructure to support high-end computing needs [19][20] - Despite the strong demand, there are concerns about the sustainability of investments in AI computing infrastructure, particularly regarding the potential for overbuilding [20][22] - The software ecosystem remains a significant challenge for domestic manufacturers, as effective software solutions are crucial for the successful deployment of high-density computing systems [18][22]

Group 1 - The core viewpoint of the report emphasizes that the era of supernodes will redefine the landscape of computing power, moving away from GPUs as the central focus to supernodes as the primary unit of computation [1][16][34] - Supernodes, which consist of multiple devices working as a single logical unit, will significantly increase the demand for advanced process technology, with the need for Scale-up switch chips expected to grow exponentially compared to traditional AI computing clusters [1][2][59] - The report highlights that the Chinese supernode market has unique opportunities, leveraging scale and energy efficiency to compensate for the performance gap with foreign counterparts, with projections indicating that by 2027, the number of domestic supernode cards will be 8.5 times that of foreign ones [3][4][30] Group 2 - The report identifies that the demand for Scale-up switches will increase nearly 40 times compared to Scale-out architectures, with specific examples such as Huawei's Atlas 950 supernode utilizing over 9,000 low-dimensional and 500 high-dimensional switch chips [2][59] - The supernode architecture is expected to revolutionize the AI computing landscape, with major players like NVIDIA, Huawei, and Alibaba already launching their supernode products, indicating a clear trend towards high-density and high-interconnectivity AI infrastructure [34][35][36] - The report outlines the advantages of supernodes in overcoming communication, power, and software bottlenecks, thus enhancing overall system efficiency and performance [26][29][59]