By Che Pan and Brenda Goh BEIJING, Dec 31 (Reuters) - China's leading maker of DRAM chips, ChangXin Memory Technologies (CXMT) Corp, said on Tuesday it plans to ​raise 29.5 billion yuan ($4.22 billion) through an initial public offering of 10.6 ‌billion shares in Shanghai. The company, which directly challenged South Korean and U.S. rivals by unveiling its latest DDR5 ‌DRAM chips last month, will use the listing proceeds to upgrade production lines and technologies, it said in a prospectus on Tuesday. ...

据介绍，Maestro是AI21的商业化支柱，年化收入约5000万美元。该平台帮助企业预处理结构化与非结 构化数据，优化输入以供AI代理高效分析，并对生成结果进行准确性验证与格式化输出。 知情人士透露，英伟达有意将Maestro整合进其企业级AI软件套件 NVIDIA AI Enterprise，以增强其在AI 代理开发与部署领域的端到端能力。此举将进一步丰富该套件中的预训练模型、开发工具及工作流管理 功能，强化英伟达在生成式AI基础设施市场的领导地位。 此次潜在收购紧随英伟达另一项重大技术合作之后——近日，该公司宣布以200亿美元获得AI芯片初创 公司Groq的技术授权，并吸纳其创始CEO及核心团队加入。Groq专注于高性能推理处理器，其技术有 望与英伟达现有GPU架构形成互补。 外媒称，AI21 Labs曾于2023年完成一轮由英伟达、谷歌、三星电子等共同参与的融资，并在今年早些 时候低调完成3亿美元新融资，估值与前轮基本持平。尽管此前已有投资关系，此次拟议收购表明英伟 达正从"财务支持者"转向"战略整合者"，意图通过垂直整合关键AI软件能力，巩固其在大模型时代软硬 协同的生态优势。（青云） 来源：环球网 ...

智通财经APP获悉，东吴证券发布研报称，2026年AI算力产业链预计将迎来业绩放量。云端算力方面， 国产GPU进入业绩兑现期，产业链协同至关重要。端侧算力则受益于AI终端创新与3D DRAM放量，多 场景加速落地。同时，先进制程晶圆代工扩产、HVDC电源架构升级以及光铜互联需求增长，共同驱动 产业链各环节技术升级与价值提升。 东吴证券主要观点如下： 云端算力：迎接国产算力产业链上下游共振带动业绩放量 3DDRAM：端侧AI存储26年为放量元年，产业趋势逐渐确立 26年AI硬件落地带来存力需求的快速提升，高带宽/低成本的3DDRAM有望在多领域放量。该行认为机 器人/AIOT/汽车等领域对本地大模型的部署离不开3DDRAM存储的支持，其为各端侧应用从"能 用"到"好用"的关键硬件革新。多款NPU的流片发布亦为3DDRAM提供丰富适配场景。此外，手机/云端 推理等场景亦将逐步导入，成为26H2及27年关键场景，应用领域持续拓宽。相关公司兆易创新、北京 君正等。 端侧AI模型：架构优化突破物理瓶颈，利益分润决定生态版图 2026年国产算力芯片龙头有望进入业绩兑现期，看好国产GPU受益于先进制程扩产带来的产能释放。考 ...

每日经济新闻 （责任编辑：董萍萍 ） 【免责声明】本文仅代表作者本人观点，与和讯网无关。和讯网站对文中陈述、观点判断保持中立，不对所包含内容 的准确性、可靠性或完整性提供任何明示或暗示的保证。请读者仅作参考，并请自行承担全部责任。邮箱： news_center@staff.hexun.com 华福证券指出，2025年第三季度半导体行业表现创历史新高，全球营收达到2163亿美元，环比增长 14.5%，首次单季突破2000亿美元大关。按照这一增速，2025年全年营收有望突破8000亿美元。半导体 行业增长动力从英伟达和存储领域向外扩散，前四大企业（英伟达、三星、SK海力和美光科技）合计 贡献了全球半导体行业超过40%的营收，凸显AI加速器和高端存储产品的持续主导地位。 半导体设备ETF（159516）跟踪的是半导体材料设备指数（931743），该指数聚焦于半导体产业链 上游的材料与设备领域，选取从事半导体材料研发生产、精密设备制造等业务的上市公司证券作为指数 样本，以反映半导体产业核心基础环节相关企业的整体表现。作为技术密集型产业的重要指标，该指数 成分股具有显著的创新属性和成长潜力，能够有效追踪全球半导体技术升 ...

投行人士认为，预先审阅机制既优化了审核资源配置效率，也强化了风险防控，体现了监管层在支 持"硬科技"与保护投资者之间的平衡。 中经记者 孙汝祥 夏欣 北京报道 上交所官网显示，长鑫科技集团股份有限公司（以下简称"长鑫科技"）科创板IPO申请，已于2025年12 月30日获受理。 公司招股说明书（申报稿）显示，长鑫科技是我国规模最大、技术最先进、布局最全的DRAM（动态随 机存取存储器）研发设计制造一体化企业。本次发行拟募资295亿元，主要用于包括DRAM存储器技术 升级项目在内的3个项目。 值得注意的是，在获受理前，长鑫科技IPO项目已完成两轮预先审阅，该项目也成为预先审阅机制出台 以来的首单IPO。 2025年6月，证监会发布科创板"1+6"改革措施，其中包括面向优质科技型企业试点IPO预先审阅机制。 7月，上交所发布施行预先审阅指引。 DRAM龙头 据长鑫科技招股说明书（申报稿）介绍，公司是我国规模最大、技术最先进、布局最全的DRAM研发设 计制造一体化企业。根据Omdia的数据，按照产能和出货量统计，公司已成为中国第一、全球第四的 DRAM厂商。 DRAM行业具有极高的技术与资金门槛，已形成一定规模的领 ...

就在本月，摩尔线程、沐曦股份先后在科创板IPO上市，在上市后，市值迅速暴涨6-7倍，甚至一度将A股的投资热情带到了一个新高度。 与此同时，同为"国产GPU四小龙"的燧原科技早在2004年就启动了科创板上市进程，另一位玩家壁仞科技也在近日向港交所提交了IPO申请、拟于2026年1 月2日在港股上市。 与此同时，诸如天数智芯、昆仑芯等更多国产AI芯片企业，也已经开启IPO征程…… 这让巴克莱银行亚太区执行主管Matt Toms也不禁感慨："如果到2026年或2027年，中国生产出低成本、有竞争力的芯片，就像当年的DeepSeek那样，我一 点也不会感到惊讶，而这将对英伟达及其供应链造成颠覆性影响。" 这就不得不让我们，在这新年旧日更替之际，重新审视以国产GPU四小龙为代表的，中国AI芯片的关键时刻。 01 学习大厂，颠覆大厂 由于国产大模型热情依旧、生成式AI已成确定性趋势，以及众所周知的原因，2025年的国产AI芯片迎来了高光时刻。 来自咨询公司Challenger, Gray&Christmas的数据显示，2025年美国因人工智能导致的裁员人数已接近5.5万人。 时至今日，生成式AI成为时代趋势，已经不容置疑 ...

Core Insights - The Singapore market presents opportunities in the small-cap sector, particularly in AI, semiconductor demand, and green electrification [1] - Focus on companies with strong balance sheets and strategic partnerships to uncover potential growth [1] Company Summaries Micro-Mechanics (Holdings) Ltd - Reported a 2.9% YoY revenue increase to S$16.7 million for 1QFY2026, driven by a 7.9% growth in the consumable tools segment [3] - Maintained a net profit margin of 18.9% despite challenges in the wafer fabrication equipment segment [4] - Generated free cash flow of S$3.8 million and holds zero debt with S$27.2 million in cash [4] - Management is optimistic with industry forecasts predicting a 26.3% surge in 2026 [4] Digital Core REIT - Owns 11 data centres with US$1.7 billion in assets under management as of December 2024 [6] - Achieved an 83.9% YoY increase in gross revenue to US$132.4 million for 9M2025 [6] - Net property income rose 49.6% to US$67.7 million, despite higher finance costs limiting distributable income growth to 1.9% [7] - Trading at a 39% discount to net asset value, indicating potential value [7] - Maintains a conservative leverage of 38.5% with room for expansion [8] CSE Global - Reported an 8.7% revenue increase to S$698.6 million for 9M 2025, driven by a 12.4% rise in the Electrification segment [10] - Strategic partnership with Amazon allows for the acquisition of over 62 million shares, enhancing its role in the data centre sector [11] - Holds an order book of S$467.5 million, positioning the company as a key infrastructure provider [12] Investment Themes - All three companies align with long-term trends in their respective sectors, showcasing resilience and strategic foresight [13] - Micro-Mechanics emphasizes precision and financial discipline, Digital Core REIT offers value in AI infrastructure, and CSE Global focuses on strategic partnerships [13]

招股书显示，此次募集资金将重点用于提升研发能力、优化高性能及超小型PA集成模组性能、高性能 滤波器研发、砷化镓外延结构及SOI工艺的开发等领域，以期突破国际巨头在BAW滤波器、毫米波模组 等领域的技术壁垒。 半导体和集成电路作为数字经济的基础要素，其战略重要性通过支持性政策、优惠税收待遇及定向投资 持续强化，推动产业向高端化、自主化升级。在政策引导与市场需求的双重驱动下，国内射频前端芯片 行业正迎来本土化浪潮。 在产品布局上，作为率先推出完全支持所有 5G 频段的 5G 射频前端解决方案的本土企业，飞骧科技已 构建完整的射频前端解决方案体系，覆盖从5G Sub-6GHz到Wi-Fi 7的前沿通信标准。其中，5G L- PAMiD模组产品在2024年已通过验证并实现量产交付，随着该产品陆续进入多家主流手机品牌旗舰机 型供应链，标志着公司在高集成度模组领域取得突破性进展。 转自：新华财经 据悉，随着募集资金的投入与核心技术的持续突破，飞骧科技有望进一步巩固全球市场地位，推动国内 射频前端产业整体升级，为我国集成电路产业实现科技自立自强、建设现代化产业体系提供坚实支撑， 成为新质生产力引领高质量发展的典范企业。（张 ...

Core Viewpoint - The AI industry is expected to transition from divergence to consensus in performance trends by 2025, with 2026 anticipated to be a significant year for domestic hard technology advancements [2] Group 1: AI Industry Trends - The AI large model's reasoning capabilities are continuously iterating, forming a closed loop with edge applications, while the demand for computing and storage hardware remains high [1] - The electronic industry's downturn since 2021 is showing signs of recovery, with a notable performance boost driven by AI innovations [2] - The global capital expenditure of the eight major cloud service providers (CSPs) is projected to grow by 40% to over $600 billion by 2026, with AI server shipments expected to increase by 20.9% [3] Group 2: Computing and Storage - Domestic chip manufacturers are actively updating their products, with Huawei planning to launch the Ascend 950PR in Q1 2026 and the Atlas 950 SuperPoD in Q4 2026 [4] - The demand for DRAM is expected to increase by 26% year-on-year in 2026, driven by AI needs, while NAND shortages are anticipated to persist, with prices rising over 40% from September to December 2025 [4] Group 3: Power and Connectivity - The optimization of data transfer and communication within servers is becoming a crucial breakthrough for enhancing computing power, with the global high-speed interconnect chip market expected to grow at a CAGR of 21.2% from 2024 to 2030 [5] - The power consumption of data center chips is projected to increase significantly, necessitating advancements in power supply architectures towards HVDC solutions [6] Group 4: AI Edge Applications - The evolution of AI from a tool to an autonomous agent capable of understanding user intent is reshaping human-computer interaction, with various consumer electronics expected to integrate these capabilities [7] - Major tech companies are expected to drive market sentiment and investment consensus through continuous exploration and innovation in AI hardware [7] Group 5: Semiconductor Industry - China's semiconductor sales are projected to account for 28% of the global market by 2024, but the domestic supply ratio remains low at 4.5% [8] - The semiconductor sector is experiencing a recovery phase, with significant growth expected in AI data centers and applications like autonomous driving and humanoid robots [8] Group 6: Key Companies to Watch - Companies such as Semiconductor Manufacturing International Corporation, Industrial Fulian, and Cambrian are highlighted as key players in the semiconductor and AI sectors [9]

Chipmaker Advanced Micro Devices Inc.’s (NASDAQ:AMD) CEO, Lisa Su, reflected on the influence of former IBM (NYSE:IBM) CEO Lou Gerstner on her early career, after the former tech executive died on Saturday, aged 83. An ‘Amazingly Curious’ Leader On Sunday, in a post on X, Su described Gerstner as a leader who was “amazingly curious and insightful” about technology, while noting that she was privileged to learn and work under his leadership early in her career. Su said she was “so honored to have had a c ...