NO.1 赛力斯向港交所递交上市申请 4月28日，港交所文件显示，赛力斯集团股份有限公司向港交所提交上市申请书。今年3月底，赛力斯就宣布拟赴港上市，其彼时表示，发行H 股股票并上市是为了深入推进全球化战略布局，打造国际化资本运作平台，提高综合竞争力。H股募资将用于新能源汽车技术研发、海外市场 拓展及供应链优化。 点评：此次赴港上市旨在打造国际化资本运作平台，提高综合竞争力。H股募资将用于新能源汽车技术研发、海外市场拓展及供应链优化，这 将有助于赛力斯在全球新能源汽车市场中占据更有利的位置。 NO.2 昭衍新药一季度业绩同比扭亏为盈 4月28日，昭衍新药（HK06127，股价15.26港元）公布2025年第一季度业绩，公司实现营业收入约2.87亿元，同比减少11.54%；归属于上市公司 股东的净利润约4111.95万元，同比扭亏为盈；基本每股收益0.05元。公告称，净利增长主要系上年同期生物资产公允价值变动损失较多所致。 点评：尽管昭衍新药一季度营业收入有所下降，但净利润实现扭亏为盈，主要得益于上年同期生物资产公允价值变动损失的减少。这表明公司 在成本控制和运营效率方面取得了一定成效。 NO.3 耀才证券金融复 ...

Core Insights - The advertising and marketing industry is undergoing transformation due to the rapid development of generative artificial intelligence (AI), with Beijing BlueFocus Data Technology Group Co., Ltd. (BlueFocus) leading the way in this shift towards an AI-driven business model [2][3]. Business Model Transformation - The current breakthroughs in AI large models are primarily in text models, while multimodal and video models are still in early stages, lacking large-scale intelligent emergence [3]. - China has made significant progress in the large model field, particularly in computational infrastructure and video models, establishing a global competitive advantage [3]. - Traditional marketing models are expected to face increasing competition and budget constraints, while companies embracing AI and accumulating data assets will differentiate themselves significantly in the next two to three years [3][4]. AI-Driven Revenue Growth - BlueFocus's "AllinAI" strategy initiated in 2023 has led to AI-enabled revenue of 108 million yuan, primarily from text content creation and data insights, which is projected to grow to 1.2 billion yuan in 2024, expanding into content generation, video production, and AI marketing automation [5][6]. - The company has deployed over 100 AI agents for various functions, with a focus on high AI concentration, high gross margins, and data accumulation for continuous AI learning [5]. Globalization Strategy - Starting in 2024, BlueFocus will implement a "Globalization 2.0 Strategy," emphasizing technology and AI-driven capabilities, with local offices established in nine countries, some of which are already profitable [6][7]. - The company aims to maintain its leading position in cross-border marketing, optimize market structure and gross margin, and expand its overseas presence by adding 3 to 5 local offices, particularly in Southeast Asia and emerging markets in Europe and South America [7].

《商道创投网》2025年4月28日从官方获悉：半导体工程智能系统企业无锡智现未来科技有限公司（以 下简称"智现未来"）近日完成了数亿元A轮融资。本轮融资由国投创业、梁溪科创母基金（博华资本） 联合领投，武汉江夏科投跟投。 《商道创投网》创业家会员·单位简介 智现未来是一家专注于半导体工程智能系统（Engineering Intelligence）的高科技企业。公司深耕半导体 制造领域20余年，凭借其多模态大模型驱动的工程智能系统解决方案，为高端制造业提供生产过程的工 程管理、产品管理以及设备智能化赋能。其核心业务涵盖设备监测、分析建模、工艺控制和良率改进等 方面，服务超过180家半导体标杆客户。公司通过"大模型+"产品应用，助力制造业转型升级，打破国际 巨头在工程智能领域的技术垄断。 《商道创投网》创业家会员·本轮融资用途是什么？ 智现未来董事长兼CEO管健博士表示，本轮融资将用于进一步强化公司在设备监测、分析建模、工艺 控制、良率改进等方面的领先优势，重点推进生成式人工智能技术在半导体制造全流程中的应用落地和 范式创新。同时，公司将助力国产半导体生态的协同优化，为客户创造更多价值。 《商道创投网》创投家会员 ...

华西证券（002926）发布研报称，在全球生成式人工智能高速发展驱动下，数据量与算力需求激增推动 光模块加速向800G/1.6T升级。国内头部光模块厂商2024年及2025Q1财报显示收入及净利润持续高增， 印证行业高景气。尽管市场担忧DeepSeek等技术降低算力成本可能抑制需求，但短期算力芯片升级、 大模型竞赛加剧供需紧张，中长期AI应用(如具身智能)仍处爆发初期，支撑光模块长期增长逻辑。叠加 地缘政治催化国产算力建设提速，行业具备业绩支撑与估值修复空间。 华西证券主要观点如下： 光模块业绩高速增长，国内云厂商算力投资加速 伴随生成式人工智能快速发展，全球数据总量与算力规模保持着高速增长态势，光模块市场向 800G/1.6T演进。 该行认为，算力需求在过去两年受益于算力芯片催生一波跳跃式需求增长，而短期供应紧缺、大厂大模 型"军备"竞赛加剧供需紧缺。中长期看AI应用仍处于极早期，包括具身智能在内的众多AI+应用场景仍 有巨大空间。作为算力集群拓展的基础组件，光模块行业仍然具备高增长能力。 短期内，考虑全球地缘政治冲突、关税贸易冲突等一系列不确定性影响下，国产算力建设加速对国内市 场需求有望提振。 光模块作 ...

Investment Rating - The report maintains a "Buy" rating for Baidu with a target price of $146.70, representing a potential upside of 61.7% from the current price of $90.74 [2][9]. Core Insights - Baidu is actively promoting its business transformation by integrating more generative AI-related search results to enhance user experience and engagement. Although the monetization of these search results has not yet begun, it is expected to start by the end of Q2, which should help boost advertising revenue growth and operating profit margins [9]. - The company has introduced several AI applications, including the Wenxin large model 4.5 Turbo and X1 Turbo, which feature multi-modal capabilities, strong reasoning, and reduced costs. The input price for the Wenxin 4.5 Turbo is only 0.8 RMB per million tokens, and the output price is 3.2 RMB, marking an 80% price reduction compared to its predecessor [9]. - Baidu's cloud business is anticipated to maintain strong revenue growth driven by increasing demand for digitalization and AI solutions, supporting overall revenue growth for the group [9]. Financial Summary - The financial forecast for Baidu shows expected sales revenue of 137,622 million RMB in FY25, with a projected growth rate of 3.4%. The adjusted net profit is expected to be 26,943.5 million RMB, reflecting a slight decrease of 0.2% [10][11]. - The company’s operating profit is projected to be 21,430 million RMB in FY25, with a slight increase in operating margin to 15.6% [11]. - Baidu's total assets are expected to grow from 427,780 million RMB in FY24 to 457,698 million RMB in FY25, indicating a healthy balance sheet [11].

Core Insights - Goldman Sachs recently engaged with Chinese software supply chain companies to discuss customer software budgets, advancements in generative AI, and growth and profitability prospects for the second half of 2025. It is anticipated that the average revenue growth for Chinese software companies covered by Goldman Sachs will be 26% year-on-year in 2025, with sequential revenue growth expected across all four quarters [1]. Group 1: Generative AI Impact - The emergence of enhanced local foundational models, such as deep search models, has shifted customer software budgets from functional tools to generative AI software, leading to increased average revenue per user and attracting new paying customers [1]. - As demand rises and profitable business models for enterprise (B-end) and consumer (C-end) users become clearer, more software companies are investing in the development of application software based on foundational models [1]. - Generative AI is enabling enterprises to reduce operational costs, such as recruitment tools, with revenue from generative AI for covered Chinese software companies expected to rise from 0%-10% in 2025 to 10%-50% by 2028 [1]. Group 2: Revenue Contributions and Projections - The contribution of generative AI revenue to various companies is projected to increase significantly by 2028, with specific companies like Kingsoft Office and iFlytek expected to see substantial growth in their generative AI revenue contributions [2]. - The revenue growth for different segments, including electronic design automation, marketing, CAD/BIM, and automotive software, is expected to be robust in 2025, with generative AI applications playing a key role [10][29]. Group 3: Software Company Strategies - Companies are shifting their business models to subscription-based services, which provide predictable revenue streams and better operational leverage, particularly in the context of generative AI applications [7]. - Software companies are focusing on providing customized solutions for large clients while also developing standardized solutions for small and medium enterprises [24][25]. - The monetization of generative AI is occurring in both B-end and C-end markets, with existing clients paying more for upgraded versions with AI features and new clients converting from free to paid status more rapidly [26]. Group 4: Growth and Profitability Outlook - Companies like Kingsoft Office and Yonyou are expected to achieve steady revenue growth, with Kingsoft focusing on expanding its AI user base and Yonyou optimizing its product mix to improve gross margins [31]. - The automotive software sector is witnessing increased demand due to the push for intelligent vehicle strategies, benefiting the entire automotive technology supply chain [29]. - Cybersecurity companies are adapting to market demands by integrating multiple point solutions into platform offerings, driven by the increasing need for security solutions in the context of generative AI [30].

"宁波膜智信息科技有限公司"为势银（TrendBank）唯一工商注册实体及收款账户 添加文末微信，加 先进封装 群 Manz亚智科技积极推进CoWoS面板化概念，即CoPoS，以解决封装效率与芯片产能问题。目 前，Manz亚智科技的FOPLP技术已步入量产阶段，同时行业也在探索基于玻璃基板的封装方 案，以进一步提升封装效能，实现更高带宽、更大密度与更强散热性能。 势银研究： 势银产业研究服务 势银数据： 势银数据产品服务 势银咨询： 势银咨询顾问服务 重要会议： 4月29日，2025势银异质异构集成封装产业大会（浙江宁波） 点此报名 当前， 随着生成式人工智能（ Generative AI）的迅猛发展，高阶AI芯片需求呈爆发式增长， 材料和封装技术迎来新的变革契机。在此背景下，突破产能瓶颈，推动先进封装成为行业新焦 点。其中，Chip-on-Wafer-on-Substrate (CoWoS) 面板化方案──Chip-on-Panel-on- Substrate (CoPoS) 概念正引领未来技术发展方向。 由此，势银（TrendBank）访谈了亚智科技事业开发部副总经理简伟铨，深入交流了当下产业 发展现状以 ...

Core Insights - The rapid development of Generative AI (GenAI) is transforming traditional cultural education, particularly in higher education, by shifting from standardized teaching to a co-creative and immersive learning model [1][10]. Group 1: Course Design Innovation - Generative AI is reshaping the teaching design logic of the "Introduction to Chinese Culture" course by embedding intelligent technology, creating a student-centered and culturally immersive learning experience [2]. - Technologies like Virtual Reality (VR) and Augmented Reality (AR) convert abstract cultural symbols into interactive experiences, enhancing students' understanding of complex cultural concepts [2][4]. - AI-driven personalized learning materials are generated based on real-time analysis of student data, allowing for dynamic optimization of course content to meet diverse student needs [3]. Group 2: Teaching Methodology Transformation - The integration of Generative AI promotes a shift from traditional teacher-led instruction to a collaborative human-machine interaction model, redefining classroom dynamics and roles [5]. - Students become active knowledge creators through AI platforms that facilitate discussions and debates, enhancing critical thinking and engagement with the material [5][6]. - Teachers transition to roles as learning experience designers and value guides, focusing on facilitating human-AI collaboration rather than solely delivering content [6][7]. Group 3: Ethical and Cultural Challenges - The application of Generative AI raises ethical concerns, particularly regarding data privacy and academic integrity, necessitating robust regulatory frameworks [8]. - There is a risk of cultural interpretation homogenization due to AI's reliance on high-frequency vocabulary, which may oversimplify complex philosophical concepts [9]. - To address these challenges, a collaborative mechanism between human experts and AI is proposed to ensure diverse cultural interpretations and maintain the depth of traditional cultural education [9][10].

Group 1: Trade Relations and Market Reactions - President Trump indicated that the cumulative 145% tariffs on Chinese goods are "too high," suggesting a significant reduction in rates, though not to zero [1] - U.S. Treasury Secretary stated that the goal of the tariff policy is not to "decouple" from China, and that negotiations may be challenging but an agreement is possible in the future [1] - U.S. stock markets rebounded with all three major indices rising over 2.5% on April 22, ending a four-day decline, driven by improved trade negotiation prospects [1] Group 2: Gold Market Dynamics - Gold prices previously surged past $3,500 per ounce but fell nearly 2% afterward, with domestic retail gold jewelry prices dropping by 27 yuan per gram [2] - The price increase in gold this year was nearly 30% due to Trump's criticisms of the Federal Reserve, making gold a primary safe-haven asset [2] - Trump's recent comments about not intending to replace Fed Chair Powell contributed to a market sentiment recovery [2] Group 3: Regulatory Challenges for Tech Giants - Apple and Meta were fined a total of €700 million by the EU under the Digital Markets Act, with Apple facing a €500 million penalty and Meta €200 million [3] - The EU's penalties are not at the maximum limit, and companies have a 60-day observation period to comply with new regulations regarding alternative purchasing methods and user consent for personalized ads [3] Group 4: Tesla's Financial Performance - Tesla reported Q1 revenue of $19.335 billion, a year-on-year decline of approximately 9%, with automotive revenue down 20% to around $14 billion [4] - Adjusted net profit fell 39% to $934 million, impacted by Model Y production upgrades, declining average selling prices, and reduced delivery volumes [4] - Tesla's energy business showed strong growth, with energy deployment increasing by 54% year-on-year [4] Group 5: Corporate Restructuring and Strategy - Intel plans to lay off over 20,000 employees, aiming to reduce bureaucracy and rebuild an engineering-focused culture [8] - This restructuring follows a previous layoff announcement of approximately 15,000 employees and is part of new CEO Pat Gelsinger's strategy to revitalize the company [8] Group 6: New Ventures in E-commerce - iQIYI has launched a content e-commerce platform focusing on live-streaming sales, targeting its membership base as a potential user group [9] - The company is seeking new growth avenues after experiencing declines in both membership and advertising revenue [9] Group 7: Financial Performance of Transsion Holdings - Transsion Holdings projected a 10.31% revenue growth for 2024, reaching 68.72 billion yuan, with African revenue accounting for 33.3% of total income [10] - Net profit is expected to be 5.55 billion yuan, a modest increase of 0.22%, influenced by market expansion and product upgrades [10] Group 8: Leadership Changes at Nike - Nike appointed Jennifer Hartley as the new Chief Strategy Officer, tasked with driving the company's strategic agenda [11] - This appointment follows a series of management changes due to performance declines, including the return of CEO Elliot Hill [11] Group 9: Automotive Innovations - Zeekr unveiled its flagship SUV, the Zeekr 9X, featuring advanced L3-level intelligent driving capabilities and a high-performance electric hybrid engine [12] - The vehicle incorporates cutting-edge technology, including a 900V architecture and multiple laser radars for enhanced driving assistance [12] Group 10: Strategic Partnerships in Automotive Technology - Intel and Black Sesame Intelligence announced a strategic partnership to develop an integrated platform for smart cockpit and autonomous driving technologies [14][15] - This collaboration aims to leverage both companies' strengths in chip technology and provide adaptable design solutions for various vehicle models [14][15]

大公司： 日产汽车下调2024财年财务展望 4月24日，日产汽车修改截至2025年3月31日的2024财年财务展望，将全年销量预测下调至335万辆，全 年销售净额预计为12.6万亿日元，营业利润预计为850亿日元，预计净亏损7000亿日元至7500亿日元。 （界面） 一季度特斯拉在欧洲销量暴跌45% 当地时间4月24日，欧洲汽车制造商协会发布的数据显示，今年前三个月，欧盟新车注册量同比下降 1.9%。电动车和混合动力车市场占比持续扩大，但传统燃油车份额大幅萎缩。数据显示，一季度，汽 油车注册量同比下降20.6%，跌至77.9万辆，占比下降至28.7%。柴油车注册量下降27.1%，市场份额降 至9.5%。纯电动汽车一季度注册量增长23.9%，达到41.3万辆，占总市场份额的15.2%，较上年同期的 12%有所上升。（央视新闻） 可口可乐将在日本涨价 可口可乐日本公司近日宣布对主要产品类别进行价格调整，自2025年10月1日起生效。可口可乐日本公 司称，在原材料和能源价格长期居高不下以及汇率波动的背景下，开展业务活动的成本持续上升。由于 市场环境的影响，仅靠企业的努力很难继续消化成本上升带来的影响。公告显示，咖啡 ...