智通财经记者 | 宋佳楠 xAI创立于2023年7月，其宗旨是了解宇宙的真实本质。自成立以来，xAI发展迅猛，推出了多个版本的Grok模型。今年 7月，xAI新一代Grok 4大模型发布，基础版已向全球用户免费开放。 值得注意的是，马斯克于10月6日回应一则关于Grok动态生成视频游戏的帖子时，确认xAI游戏工作室将于2026年底前 发布一款AI生成游戏，表明xAI的产品覆盖范围将进一步延伸。 参与投资xAI的英伟达近期频繁出手。9月18日，英伟达与英特尔宣布达成 "历史性合作"。英伟达以每股23.28美元的价 格收购英特尔普通股，总投资额达50亿美元，持有英特尔超过4%的股权。 双方将在数据中心领域联合开发人工智能系统，英特尔将专门为英伟达定制x86架构CPU，这些处理器将与英伟达GPU 深度集成，形成高性能计算节点。在个人计算领域，英特尔还将生产集成英伟达RTX GPU的系统级芯片。 9月22日，英伟达又宣布将向OpenAI投资最多1000亿美元，称将助力OpenAI构建和部署至少10吉瓦的AI数据中心，这些 数据中心包含数百万块英伟达GPU。首个吉瓦级算力集群将基于英伟达下一代Vera Rubin平台搭 ...

记者丨 倪雨晴 编辑丨张星 没有永远的敌人，芯片巨头英伟达和英特尔走到了一起。 竞争激烈的半导体市场上，从来都是兼并收购、投资重整的历史。 这一次，全球GPU和CPU龙头结成了历史性的联盟。 9月18日，英伟达与英特尔宣布达成合作，将共同开发AI基础设施及个人计算产品。同时，英伟达将以每股23.28美元的价格，向英特尔普通股 投资50亿美元。 一夜之间，英特尔迎来了久违的"高光时刻"。当天美股开盘后，英特尔股价瞬间拉升近30%，截至最新收盘，英特尔市值达到1295亿美元，今 年涨幅47.53%。 这背后，一方面的推力来自英伟达的50亿美元真金白银；另一方面，芯片产品的合作，为双方打开增量市场，尤其对于英特尔而言，直接登上 了AI大船。 而AI算力市场的棋盘发生微妙变化，AMD、ARM、台积电等竞合复杂的老牌竞争对手们，都将面临新的格局。 此外，值得注意的是， 美国政府8月刚刚入股英特尔，谈及和此番交易的关系，英伟达创始人兼CEO黄仁勋表示："我们认为，这将是一笔不 可思议的投资……特朗普政府完全没有参与这项合作，当然，他们应该也会非常支持。"他强调，合作只关乎技术与产品。 拿到A I船票 英伟达以50亿美元直 ...

Core Viewpoint - The collaboration between Nvidia and Intel is seen as a potential game-changer in the semiconductor industry, combining Nvidia's AI chip capabilities with Intel's extensive user base and patents, targeting a market exceeding $50 billion [1][3]. Group 1: Collaboration Announcement - Nvidia and Intel announced their partnership on September 18, focusing on creating innovative solutions for the next computing era, particularly in data centers, personal computers, and edge computing [3]. - The collaboration includes Nvidia acquiring a stake in Intel at $23.28 per share, and both companies will develop customized data center and PC products together [5][6]. Group 2: Market Impact - Following the announcement, Nvidia's stock rose over 3%, while Intel's surged more than 20%, indicating strong market optimism about the partnership [3]. - The collaboration is expected to disrupt the current market dynamics, potentially ending the existing three-way competition among Nvidia, Intel, and AMD [9]. Group 3: Strategic Opportunities - Nvidia can now integrate Intel's x86 architecture CPUs into its ecosystem, creating new types of integrated graphics laptops [8]. - The partnership allows both companies to become major customers for each other, enhancing revenue growth through mutual sales of CPUs and GPUs [8]. Group 4: Industry Dynamics - Analysts suggest that the collaboration may alter the competitive landscape, making it difficult for AMD to compete against the combined strengths of Nvidia and Intel [9][10]. - The partnership reflects a strategic restructuring in the U.S. semiconductor industry, reinforcing its global advantage in the semiconductor and AI sectors [14][15]. Group 5: Future Considerations - The collaboration is viewed as a trial partnership, with uncertainties regarding the extent of technological integration and cooperation between the two companies [11]. - Intel's vast patent portfolio may provide long-term advantages, potentially shifting the balance of power in the partnership as it recovers from recent challenges [15].

Core Viewpoint - The article discusses the strategic investment and partnership between Nvidia and Intel, highlighting the implications for the semiconductor industry and the companies involved [2][5]. Group 1: Investment Details - Nvidia announced a $5 billion investment to acquire Intel shares at $23.28 per share, making it a significant shareholder in Intel [2][5]. - The investment price represents a discount of approximately 6.5% compared to Intel's closing price of $24.9 on the previous Wednesday, but is higher than the U.S. government's recent investment price of $20.47 per share [5]. Group 2: Strategic Partnership - Nvidia and Intel will collaborate on data center and personal computing products, with Intel customizing x86 architecture CPUs for Nvidia's AI infrastructure [5][6]. - Intel will also produce x86 system-on-chip (SoC) solutions that integrate Nvidia's RTX GPUs to meet high-performance computing and AI terminal demands [5][6]. Group 3: Market Reaction - Following the announcement, Intel's stock price surged over 30% in pre-market trading, while Nvidia's shares increased by over 3% [5].

Core Viewpoint - Nvidia announced a $5 billion investment to acquire Intel's common stock at $23.28 per share, pending regulatory approval, aiming to enhance collaboration in developing customized data center and PC products [3][4]. Group 1: Company Collaboration - Nvidia and Intel will collaborate to develop multiple generations of customized data center and PC products, focusing on large-scale enterprise and consumer market applications [4]. - Intel will customize x86 architecture CPUs for Nvidia, which will integrate these products into its AI infrastructure platform [4]. - In the personal computing sector, Intel will develop x86 architecture SoCs integrated with Nvidia's RTX GPU chips to drive computers [4]. Group 2: Technological Integration - The collaboration will utilize Nvidia's NVLink to connect Nvidia and Intel architectures, combining Nvidia's AI and accelerated computing advantages with Intel's CPU technology and x86 ecosystem [4]. - Nvidia's CEO Jensen Huang emphasized that this partnership will tightly integrate Nvidia's AI and accelerated computing stack with Intel's CPU and x86 ecosystem, expanding the ecosystem for both companies [4][5]. Group 3: Market Reaction - Following the announcement, Intel's stock rose by 28.39%, while Nvidia's stock increased by 3.3% [5]. - As of September 17, Intel's stock closed at $24.9 per share, with a market capitalization of $109 billion, while Nvidia's market capitalization stood at $4.14 trillion [5].

Core Viewpoint - The recent acceptance of IPO applications for four semiconductor companies on the Sci-Tech Innovation Board reflects a complete industrial chain layout from "materials - equipment - design," highlighting the urgent need for domestic semiconductor industry autonomy and the capital market's support for the sector [1][10]. Group 1: Company Summaries - **Zhaoxin Integrated**: A domestic CPU manufacturer focusing on x86 architecture, with products for desktops, servers, and workstations. The company aims to raise 4.169 billion yuan for new processor development and capacity building, having been accepted for IPO on June 17 [2][3]. It is currently unprofitable but expects to turn a profit by 2027 [3][4]. - **Chipmi Technology**: Specializes in the development and production of semiconductor-grade perfluoroether rubber seals, crucial for etching and thin-film deposition processes. The company plans to raise 785 million yuan for seal development and industrialization, having been accepted for IPO on June 16 [5][6]. Revenue is projected to grow from 41.59 million yuan in 2022 to 208 million yuan in 2024, with a significant increase in net profit and gross margin [6]. - **Shanghai Super Silicon**: A leading manufacturer of semiconductor wafers, producing 300mm and 200mm silicon wafers for memory and logic chips. The company aims to raise 4.965 billion yuan for expanding 300mm silicon wafer production and high-end material research, accepted for IPO on June 13 [8][9]. Despite revenue growth, it has faced continuous losses due to underutilization of production capacity [8]. - **Hengyun Chang**: Focuses on core components for semiconductor equipment, particularly plasma RF power systems, breaking the monopoly of American companies. The company plans to raise 1.55 billion yuan for industrialization and R&D center construction, accepted for IPO on June 13 [10]. Revenue is expected to increase from 158 million yuan in 2022 to 541 million yuan in 2024, with a leading market share in domestic plasma RF power systems [10]. Group 2: Industry Insights - The acceptance of these four companies' IPOs indicates a strong focus on supporting the semiconductor industry, which is crucial for achieving self-sufficiency in the domestic supply chain [10]. The companies cover key segments of the semiconductor industry, from design to materials and equipment, addressing critical technology gaps [10].