Clear Profit Effect: Multiple Funds Snag Cornerstone Investments 赚钱效应明显，多路资金抢筹基石投资

Summary

The Hong Kong stock new share market is showing significant profitability effects, with 81.32% of the 91 IPOs this year introducing cornerstone investors, attracting long-term capital such as foreign investment and insurance funds to compete for positions.
Embodied intelligence infrastructure company "Lingjing Zhiyuan" has completed its angel round and angel+ round financing totaling over 100 million RMB, led by Matrix Partners, with funds allocated to chip R&D and ecosystem building.
The world’s first ultra-high frame rate world model has been created, achieving 50 frames per second without NVIDIA hardware, marking a breakthrough for non-NVIDIA architectures in video generation.
Major tech companies collectively removing AI chatbots reflects corporate strategies to avoid compliance risks amid tightening regulations.

Deep Analysis

TL;DR

• The Hong Kong stock new share market is showing significant profitability effects, with 81.32% of the 91 IPOs this year introducing cornerstone investors, attracting long-term capital such as foreign investment and insurance funds to compete for positions.
• Embodied intelligence infrastructure company "Lingjing Zhiyuan" has completed its angel round and angel+ round financing totaling over 100 million RMB, led by Matrix Partners, with funds allocated to chip R&D and ecosystem building.
• The world’s first ultra-high frame rate world model has been created, achieving 50 frames per second without NVIDIA hardware, marking a breakthrough for non-NVIDIA architectures in video generation.
• Major tech companies collectively removing AI chatbots reflects corporate strategies to avoid compliance risks amid tightening regulations.

Why It’s Worth Reading

This article covers capital market allocation trends toward AI and hard tech assets (Hong Kong cornerstone investments), the latest financing progress in the embodied intelligence sector, and breakthroughs in underlying technological architectures. It provides a multi-dimensional perspective for understanding fund flows and technological evolution within the AI industry chain. For professionals tracking primary market investment trends and secondary market tech stock performance, these insights help grasp key nodes in industry cycles and technological iterations.

Technical Analysis

• Hong Kong Cornerstone Investment Mechanism: Data shows that long-term capital (foreign investors, insurance funds, bank wealth management subsidiaries) serves as a crucial support for Hong Kong new shares due to their long liability durations and large scale. This "safety cushion" effect not only enhances post-listing stock price stability but also reflects market recognition of the long-term allocation value of high-quality Hong Kong assets.
• Use of Funds by Lingjing Zhiyuan: The company focuses on embodied intelligence infrastructure, with financing priorities centered on "embodied native chip R&D." This indicates that the industry is moving from the application layer deeper into customized underlying hardware and standardized ecosystem construction, aiming to meet specific demands for dedicated computing power and energy efficiency in embodied intelligence.
• Breakthrough in Ultra-High Frame Rate World Model Technology: The newly created world model achieves high-frame-rate generation at 50 fps and is explicitly noted as having "0% NVIDIA content." This suggests the adoption of alternative hardware acceleration solutions (such as AMD, Intel, or custom ASICs) or software stack optimizations, demonstrating the technical feasibility of reducing dependence on NVIDIA and holding symbolic significance for diversifying the computing power supply chain.

Industry Implications

• Reshaping Valuation Logic for AI Assets: With deep involvement of long-term capital in Hong Kong AI-related IPOs, market pricing mechanisms are shifting from short-term concept speculation to long-term fundamentals and cash flow expectations. Investors need to focus more on enterprises' sustainable profitability and asset quality.
• Embodied Intelligence Enters a Period of Infrastructure Deepening: Financing activities by companies like Lingjing Zhiyuan indicate that industry focus has shifted from algorithmic models to underlying chips and standard ecosystems. Startups must build barriers in hardware adaptation and industry standard-setting rather than remaining solely at the application layer.
• Dual Drivers of Compliance and Technological Self-Reliance: The removal of AI chatbots by major firms highlights compliance pressures, while the emergence of an "NVIDIA-free" world model reflects the trend toward technological self-reliance. Future AI development will proceed along two dimensions: strict regulation and supply chain diversification.

TL;DR

• 港股新股市场呈现显著赚钱效应，年内91家IPO中81.32%引入基石投资者，吸引外资、险资等长线资金竞相布局。
• 具身智能基础设施公司“灵境智源”完成超亿元人民币天使轮及天使+轮融资，由经纬创投领投，资金用于芯片研发及生态构建。
• 全球首个超高帧世界模型诞生，实现50帧运行且不含英伟达硬件，标志着非NVIDIA架构在视频生成领域的突破。
• 大厂集体下架AI聊天机器人，反映出监管趋严背景下企业对合规风险的规避策略。

为什么值得看

本文涵盖了资本市场对AI及硬科技资产的配置动向（港股基石投资）、具身智能赛道的最新融资进展以及底层技术架构的创新突破，为理解AI产业链的资金流向和技术演进提供了多维视角。对于关注一级市场投融资趋势及二级市场科技股表现的从业者而言，这些信息有助于把握行业冷暖与技术迭代的关键节点。

技术解析

• 港股基石投资机制：数据显示，长线资金（外资、险资、银行理财子）因负债端久期长、体量大，成为港股新股的重要支撑。这种“安全垫”效应不仅提升了新股上市后的股价稳定性，也反映了市场对港股优质资产长期配置价值的认可。
• 灵境智源融资用途：该公司专注于具身智能基础设施，融资重点在于“具身原生芯片研发”。这表明行业正从单纯的应用层向底层硬件定制化和标准化生态构建深入，旨在解决具身智能对专用算力和能效比的特定需求。
• 超高帧世界模型技术突破：新诞生的世界模型实现了50帧的高帧率生成，且明确标注“英伟达含量0”。这暗示其可能采用了不同的硬件加速方案（如AMD、Intel或自研ASIC）或软件栈优化，证明了去NVIDIA依赖的技术可行性，对算力供应链多元化具有标志性意义。

行业启示

• AI资产估值逻辑重塑：随着长线资金深度介入港股AI相关IPO，市场定价机制正从短期概念炒作转向长期基本面与现金流预期，投资者需更关注企业的可持续盈利能力和资产质量。
• 具身智能进入基建深耕期：灵境智源等公司的融资动向显示，行业焦点已从算法模型转向底层芯片与标准生态，初创企业需在硬件适配和行业标准制定上建立壁垒，而非仅停留在应用层。
• 合规与技术自主可控双重驱动：大厂下架AI聊天机器人凸显合规压力，而“去英伟达”世界模型的诞生则体现了技术自主可控的趋势。未来AI发展将在严格监管与供应链多元化两个维度下并行推进。