Core Viewpoint - The recent announcements from Magmi Tech indicate significant changes in shareholder composition, particularly the reduction of shares held by associated parties of Ge Weidong, which may impact the company's stock performance and investor sentiment [1][2]. Shareholder Changes - As of January 26, Ge Weidong's associate, Ge Guilian, has exited the top ten shareholders list after reducing her holdings by at least 760,000 shares since the beginning of the year [1][2]. - Ge Guilian held 606.76 million shares as of the end of 2025, while another associate, Wang Ping, maintained her holdings at 11.49 million shares, unchanged from the previous year [2][3]. Stock Performance - Magmi Tech's stock price has seen a dramatic increase, rising from below 18 yuan at the beginning of 2024 to over 140 yuan recently, marking a peak increase of over six times [1][7]. - However, the stock has experienced a decline of over 15% since January 29, indicating volatility in the market [7]. Financial Performance - The company has projected a net profit of 120 million to 150 million yuan for 2025, representing a year-on-year decline of 65.61% to 72.48% due to increased cost pressures and rising operational expenses [9]. - The decrease in gross margin is attributed to the company's extensive product lines and the cost pressures faced as a component supplier [9]. AI Business Outlook - Magmi Tech has begun receiving small batch orders for AI server power supplies, which are expected to contribute significantly to revenue as production scales up [10]. - Despite recent market fluctuations, the AI sector remains a key investment focus, with expectations for continued growth in technology and new productivity sectors [10][11].

作为国家部委、河南省、郑州市等多方大力支持下重点部署的全国一体化算力网络重大基础设施，核心节点的战略意义毋庸置疑。此次开创性完成 国产万卡超集群部署，不仅实现多项业界之最，也为中国算力网络建设提供一个国家级样板间。 集成度最高 ： scaleX 万卡超集群采用了全球首创高密度单机柜，基于超高密度刀片、浸没相变液冷等技术，单机柜算力密度大幅提升 20 倍， PUE 值低至 1.04 。 集成规模最大 ： scaleX 是国内首个万卡超集群，基于 scaleFabric 高速网络可实现 10240 块加速卡集成部署，大集群合体同样达到国 产计算集群最大规模。 开放性最强 ： 该集群完全基于 AI 计算开放架构设计，打破了传统的单一技术栈闭环依赖，可以支持多品牌 AI 加速卡，并兼容 AI 计算主 流软件生态，产业上下游接入能力极强。 众所周知，算力网络的建设目标是为了打破算力基础设施异构壁垒，解决国内算力资源分布不均、利用率不足等问题。 scaleX 万卡超集群 基于开放架构实现业界单体最大规模算力供给，无疑为算力基础设施建设提供了可复制、验证和普及落地的样板。 首先，开放架构体系让异构算力互联的问题不攻自破。 ...

Core Insights - The firewall market in China is steadily growing, with the market size expected to increase from $1.17 billion in 2016 to $3.02 billion by 2025, and projected to reach $3.156 billion in 2026 [2][7][20] - The growth is driven by policy support, increasing market demand, and technological advancements, particularly with the deployment of 5G, private networks, and industrial internet [2][15] - The rise of edge computing is increasing the demand for lightweight firewalls, and customized firewall products for scenarios like IoT and smart cities are expected to emerge [2][15] Firewall Industry Definition and Classification - A firewall is an access control product that creates barriers between internal networks and insecure external networks, preventing unauthorized access and attacks [3][16] - Firewalls can be categorized based on technology into traditional firewalls, Unified Threat Management (UTM) firewalls, Next-Generation Firewalls (NGFW), and AI firewalls [5][18] - They can also be classified by form into software firewalls, hardware firewalls, and chip-level firewalls [5][18] Current Development Status - The firewall industry is characterized as a critical component of IT infrastructure, acting as a gatekeeper for network security [7][20] - The market is experiencing steady growth due to supportive policies and increasing technological advancements [7][20] Industry Value Chain - The upstream of the firewall industry includes basic hardware (chips, hard drives, network cards, etc.) and basic software (BIOS, operating systems, encryption) [9][22] - The midstream consists of firewall products, while the downstream applications span various sectors including government, finance, telecommunications, energy, education, healthcare, transportation, and aerospace [9][22] Key Companies - Notable listed companies in the firewall sector include Sangfor Technologies (300454.SZ), Tianrongxin (002212.SZ), Hillstone Networks (688030.SH), Deepin Technology (300768.SZ), Ruijie Networks (301165.SZ), Venustech (002439.SZ), Qihoo 360 (688561.SH), and NSFOCUS (300369.SZ) [2][15]

（来源：科技旋涡） 国家超算互联网迈入规模化新阶段 近日，国家超算互联网应用技术大会暨核心节点上线试运行仪式在郑州举行。 据悉，该节点算力资源由曙光scaleX万卡超集群系统提供，可对外输出超3万卡的国产AI算力，这也是国家超算互联网平台上线以来接入的全国最大单体 国产AI算力资源池。 相较于算力规模体量的突破，该节点带来的超算互联网运营模式更让AI用户直接受益。 国家高性能计算机工程技术研究中心副主任曹振南介绍称，国家超算互联网平台旨在为各类前沿应用场景带来高效普惠的计算服务，让算力像水电一样自 由流通、高效易用。 业内有媒体指出，近年来，国产AI计算产业面临两个极端： · 一边是AI模型需求爆炸式增长 基于此，该节点采用的曙光scaleX万卡超集群完全基于AI计算开放架构设计，可全面兼容CUDA等主流软件生态，支持多品牌国产加速卡混合部署。 这也让其不仅具备了向十万卡、百万卡规模的灵活扩展能，还能广泛接入上下游，实现一体化算力统筹调度。 另外，国家超算互联网平台通过"算力+应用"一体化服务，正在向类京东、淘宝模式的算力应用商城模式迈进。 数据显示，截至2025年底： 大幅降低了算力成本和使用门槛。 这一模 ...

面对行业算力供给不足、供需脱节、生态不兼容等痛点，该节点搭载的曙光scaleX万卡超集群基于AI计 算开放架构，兼容主流生态、支持多品牌国产加速卡混合部署，已完成400多个主流大模型适配优化， 实现"开箱即用"，大幅降低算力使用门槛；同时，依托国家超算互联网实现一体化算力调度，可接入上 千款应用，为全球用户提供好用易用的中国AI算力。 经济观察网2月5日，国家超算互联网核心节点在郑州上线试运行，该节点采用中科曙光scaleX万卡超集 群，可提供超3万卡国产AI算力，是国家超算互联网平台接入的全国最大单体国产AI算力资源池，可支 撑万亿参数大模型训练、高通量推理、AI for Science等大规模AI计算场景。 ...

贾跃亭发布人形机器人 北京时间2月5日早上7:30，FF在拉斯维加斯举办美国国家汽车经销商大会活动上，举行FF首批具身智能（Embodied AI）机器人产品发布会。此 前，FF从上市到几近破产，去年4月，贾跃亭重新出任FF的Co-CEO，试图重塑FF。身背巨额债务的贾跃亭，自2017年7月赴美创业，已八年未回 国。 黄仁勋驳斥AI引发的软件股抛售潮为"世界上最不合逻辑的事情" 英伟达首席执行官黄仁勋表示，本周市场因担心人工智能带来的颠覆性影响而抛售软件股，这毫无道理。英伟达的设备位于全球人工智能数据 中心建设的核心。黄仁勋在思科周二晚些时候的一场活动中表示，软件产品是工具，人工智能会使用这些工具，而不是重新发明它们。 中国最大国产AI算力池上线 贾跃亭称，FF发布三大系列EAI机器人：Futurist系列是全尺寸职业型具身智能人形机器人；Master系列是运动型具身智能人型机器人；Aegis系 列是安防和陪伴型专业四足具身智能机器人。发文配图显示，Futurist系列机器人定价34990美元起，Master系列机器人定价19990美元起，Aegis 系列机器人定价2499美元起。 马云现身阿里总部千问春节项 ...

北京商报讯2月5日，中科曙光宣布，由公司提供的3套万卡超集群系统在国家超算互联网郑州核心节点 同时上线试运行，成为全国首个实现3万卡部署、且实际投入运营的最大国产AI算力池，全面覆盖万亿 参数模型训练、高通量推理、AI for Science等大规模AI计算场景。 (相关资料图) ...

北京商报讯（记者 陶凤 王天逸）2月5日，中科曙光宣布，由公司提供的3套万卡超集群系统在国家超算 互联网郑州核心节点同时上线试运行，成为全国首个实现3万卡部署、且实际投入运营的最大国产AI算 力池，全面覆盖万亿参数模型训练、高通量推理、AI for Science等大规模AI计算场景。 ...

Core Viewpoint - Starting from 2025, supernodes will become a significant technological innovation direction in the AI computing network, with increasing competition among AI chip manufacturers in both chip performance and Scale up network [1][5]. Group 1: Supernode Development - Nvidia has launched mature supernode solutions, with plans to release GH200 NVL72, GB200/GB300 NVL72, and VR200 NVL72 from 2024 to 2026 [1][3]. - The Blackwell architecture standardizes Scale up with GB200 NVL72 stabilizing the scale at 72 GPUs per cabinet, consisting of 18 Compute Trays and 9 Switch Trays [2]. - The Rubin architecture will enhance bandwidth, with the NVLink 6 Switch increasing single GPU interconnect bandwidth to 3.6 TB/s, up from 1.8 TB/s [2]. Group 2: Nvidia's Competitive Advantage - Nvidia maintains a leading position in the supernode market, with a projected shipment of approximately 2,800 units of GB200/300 NVL72 by 2025 [3]. - Future plans include the introduction of Vera Rubin NVL144 and Rubin Ultra NVL576, expanding interconnected GPUs from 72 to 576 [3]. - Innovations such as NVLink and NVLink Switch are crucial for achieving high bandwidth and low latency in AI training clusters, with NVLink 5 Switch supporting a total bandwidth of 130 TB/s for 72 GPUs [4]. Group 3: Industry Landscape and Investment Strategy - The global supernode competition landscape is still being established, with Nvidia currently in a leading position [6]. - The report suggests monitoring Nvidia's supernode supply chain, including components like PCB backplanes, high-speed copper cables, optical modules, and cooling systems [6]. - Chinese manufacturers are actively participating in the supernode and Scale up network sectors, with potential for domestic firms to gain a competitive edge [6].

Group 1 - The core viewpoint of the news highlights the fluctuations in the AI industry chain, with a focus on the performance of the CSI Artificial Intelligence Theme Index and the AI Artificial Intelligence ETF, which has seen significant net inflows recently [1][2] - As of February 5, 2026, the CSI Artificial Intelligence Theme Index (930713) showed mixed performance among its constituent stocks, with Stone Technology leading at a 2.78% increase, while the AI Artificial Intelligence ETF (512930) experienced a decline of 1.78% [1] - The AI Artificial Intelligence ETF has recorded a total net inflow of 209 million yuan over the past three days, with a peak single-day net inflow of 12.8 million yuan, indicating strong investor interest despite recent market adjustments [1] Group 2 - The top ten weighted stocks in the CSI Artificial Intelligence Theme Index account for 57.27% of the index, with notable companies including Zhongji Xuchuang, Xinyi Sheng, and Hanwha Technology [2] - The AI Artificial Intelligence ETF closely tracks the CSI Artificial Intelligence Theme Index, which comprises 50 listed companies involved in providing foundational resources, technology, and application support for artificial intelligence [1][2]