A股市场上或将迎来一家算力巨头。 1月6日，证监会官网显示，超聚变数字技术股份有限公司（以下简称"超聚变"）启动上市辅导。 脱胎于华为的超聚变，目前由河南国资控股，中国移动、中国电信以及中金资本、招商资本、高瓴资本 等现身股东榜，股东阵容豪华。 超聚变实力不菲。官网显示，公司在全球部署10个研发中心与6个供应中心，服务全球100多个国家和地 区的一万多家客户，拥有核心专利超2000件。 公开信息显示，2024年，超聚变销售收入突破400亿元，2025年的目标是突破500亿元。 受益于超聚变筹划IPO消息，1月7日，参投超聚变的东方明珠股票涨停。与此同时，此前曾频频被传是 超聚变借壳上市对象的荣科科技、安彩高科等超聚变概念股大跌。 长江商报消息 ●长江商报记者 沈右荣 如果超聚变顺利登陆A股市场，将会带给市场什么样的想象空间？ 大量资本提前押注 传闻已久的借壳上市没有照进现实，超聚变启动上市辅导，加速冲击IPO上市。 证监会官网显示，1月6日，超聚变启动上市辅导，辅导机构为中信证券，辅导协议签署时间为2025年12 月31日。 根据上市辅导备案报告，超聚变成立于2021年9月13日，注册资本约为8.8亿元，法定 ...

据证监会官网显示， 超聚变数字技术股份有限公司于2026年1月6日正式启动上市辅导，中信证券担任辅导机构。 超聚变前身是华为X86服务器业务，2021年，在海外因素的推动下，华为作出了重大战略调整，将旗下该业务以超聚变为新实体进行独立 运营。 营收方面，2024年公司销售收入突破400亿元。2025年初，超聚变董事长刘宏云透露，一季度该公司营收再翻倍，2025全年营收目标将突 破500亿元。 产品方面，超聚变的服务器产品体系，核心承接了原华为FusionServer系列与KunLun（昆仑）系列的技术积淀，进一步细化为通用服务 器、人工智能服务器及关键业务服务器三大核心品类。 值得关注的是，在2025年世界人工智能大会上，超聚变加码智算领域，展台上发布多元智算即插即用超级集群系统。 ...

Group 1 - Goldman Sachs has updated its global server market forecast, expecting total revenue to reach $433.1 billion, $606.1 billion, and $763.9 billion in 2025, 2026, and 2027 respectively, with year-on-year growth rates of 71%, 40%, and 26% [2] - AI training servers are identified as the core growth engine, with projected revenues of $234.8 billion, $369.8 billion, and $506.2 billion for the same years, reflecting year-on-year growth rates of 97%, 57%, and 37% [2] - The report highlights a structural transformation in the global server market, driven by accelerated ASIC server penetration and significant capital expenditure growth from global cloud service providers, maintaining a high prosperity period from 2025 to 2027 [1][2] Group 2 - ASIC chip penetration in AI servers is expected to increase, with forecasts of 38%, 40%, and 50% for 2025, 2026, and 2027 respectively, up from a previous estimate of 45% for 2027 [3] - The demand for AI chips corresponding to AI servers is projected to reach 11 million, 16 million, and 21 million units in 2025, 2026, and 2027, representing increases of 7%, 17%, and 26% from previous forecasts [3] - The AI rack server market is shifting from reliance on Nvidia to a more diversified competition, with non-Nvidia solutions like AMD's Helios expected to gain market share [4] Group 3 - High-power AI training servers are projected to see significant growth, with shipment forecasts of 692,000, 952,000, and 1,227,000 units for 2025, 2026, and 2027, and corresponding market sizes of $180.2 billion, $205.2 billion, and $251.1 billion [5] - AI inference servers are expected to grow steadily, with shipment forecasts of 470,000, 539,000, and 656,000 units, and market size increasing from $29.8 billion to $48.4 billion from 2025 to 2027 [6] - The general server market is returning to normal growth, with shipment growth rates of 11%, 8%, and 2% for 2025, 2026, and 2027, and revenue growth rates of 51%, 19%, and 5% [7] Group 4 - Key companies in the server supply chain include ODM manufacturers like Wistron, Quanta, and Hon Hai, with Hon Hai being a leader in AI server market share [8] - Liquid cooling manufacturers such as AVC and Auras are highlighted for their roles in the cooling solutions for AI servers, with AVC providing custom cooling solutions for Nvidia's platforms [10][11] - TSMC is recognized as a foundational player in the AI chip and ASIC manufacturing sector, while companies like Chenbro and GCE are noted for their roles in critical components for server manufacturing [12]

Core Viewpoint - The company, Gongjin Co., Ltd. (603118), has reported an increase in production and revenue from general and AI servers for domestic clients in 2023, and is exploring new business opportunities in the EMS product line, including robotic vacuum cleaners and pool cleaning robots by 2025 [1] Group 1 - The company primarily engages in contract manufacturing of servers for domestic clients [1] - There has been an increase in both the quantity and value of general and AI servers produced for clients this year [1] - The company is planning to introduce new business segments, specifically robotic vacuum cleaners and pool cleaning robots, into its EMS product line by 2025 [1] Group 2 - Specific details regarding client collaborations are not disclosed due to commercial agreements [1]

Core Viewpoint - The company has seen an increase in the production quantity and value of general and AI servers for domestic clients this year [1] Group 1 - The company primarily engages in contract manufacturing of servers for domestic clients [1] - The production output and value scale of general and AI servers have both improved this year [1] Group 2 - In 2025, the company plans to explore new business growth avenues by introducing EMS product lines, including robotic vacuum cleaners and pool cleaning robots [1]

Core Viewpoint - The company has seen an increase in the production and output value of general and AI servers for domestic clients in 2023, while also exploring new business avenues for growth by introducing EMS product lines such as robotic vacuum cleaners and pool cleaning robots [1] Group 1 - The company primarily produces servers for domestic clients, with a focus on contract manufacturing [1] - There has been an increase in both the quantity and output value of general and AI servers provided to clients this year [1] - The company is actively seeking new growth opportunities by introducing new EMS product lines, including robotic vacuum cleaners and pool cleaning robots, by 2025 [1] Group 2 - Specific details regarding collaborations with clients, including the AI terminal testing products developed in partnership with ByteDance, cannot be disclosed due to commercial agreements [1]

Core Insights - The company, Huqin Technology, is expected to achieve significant growth in AI servers and general servers, as well as switches by 2025 [1] - The data center business is projected to maintain a doubling growth in revenue for the first three quarters of 2025, with an estimated annual revenue exceeding 40 billion yuan [1] Group 1 - The company anticipates a multiple growth in AI server segment by 2025 [1] - The company also expects rapid growth in general servers and switches [1] - The overall revenue for the data center business is projected to exceed 40 billion yuan for the entire year [1]

Core Viewpoint - Company is a leading global intelligent hardware platform enterprise focused on the development and application of artificial intelligence technology, continuously investing in AI-related product layout [1] Group 1: AI and Data Business Strategy - Company emphasizes a dual-driven strategy of "CSP customers + industry channel market" to enhance cooperation efficiency and deepen relationships with clients in the data business sector [1] - Company has achieved large-scale shipments of full-stack products including AI servers, general servers, and switches for top domestic internet clients [1] Group 2: R&D and Manufacturing Capabilities - Company has established strong R&D and manufacturing capabilities in the data business, securing a core supplier position among major CSP clients [1] - Company has conducted extensive R&D, adaptation, and testing to meet the future potential demands of mainstream clients across various computing power platforms and product forms [1] Group 3: Future Growth Expectations - Company anticipates continued revenue growth in the data business, driven by the booming development of artificial intelligence [1] - Company aims to enhance the technical R&D and delivery capabilities of various AI product terminals, striving for more innovations and breakthroughs [1]

Core Viewpoint - Company is a leading global intelligent hardware platform enterprise focusing on the development and application of artificial intelligence technology, continuously investing in AI-related product layout [1] Group 1: AI and Data Business - Company emphasizes deepening cooperation with clients in the data business sector to enhance collaboration efficiency [1] - The company has implemented a dual-driven strategy of "CSP clients + industry channel market," achieving large-scale shipments of full-stack products including AI servers, general servers, and switches for top domestic internet clients [1] - Company has established a strong R&D and manufacturing capability in the data business, securing a core supplier position among major CSP clients [1] Group 2: Future Outlook - Company anticipates continued revenue growth in the data business, driven by the booming development of artificial intelligence [1] - The company aims to enhance the technical R&D and delivery capabilities of various AI product terminals, striving for more innovations and breakthroughs [1]

Group 1 - Shuaifeng Electric announced an investment of 53 million yuan as a limited partner in the Xiamen Chip Force Lan Fund, which will directly invest in Super Fusion Digital Technology Co., Ltd. This investment represents about half of the fund's total subscribed capital of 105.838 million yuan [2][4] - This marks the second listed company to indirectly invest in Super Fusion within a month, alongside other institutions and individual investors such as Advantage Jin Control (Shanghai) Asset Management Co., Ltd. and Fujian Bingzhong Investment Co., Ltd. [4] - Super Fusion Digital Technology Co., Ltd. was formerly Huawei's X86 server business, which was spun off in 2021 and is now a leader in AI servers in China, with a product line that includes general servers, AI servers, and critical business servers [4] Group 2 - Shuaifeng Electric's core integrated stove business is facing challenges due to shrinking market demand and intense competition, leading to a decline in performance. The company's revenue for the first three quarters was 190 million yuan, a year-on-year decrease of 39.42%, and the net profit attributable to shareholders was 2.9455 million yuan, down 93.32% [4] - In the third quarter alone, Shuaifeng Electric reported revenue of 59.39 million yuan, a year-on-year decline of 25.6%, and a net loss of 350,000 yuan, a decrease of 109.1% year-on-year [4] - The investment aims to achieve financial returns and allows Shuaifeng Electric to enter the rapidly developing computing infrastructure sector through the fund subscription [4]