作者 林更 2026年1月，拉斯维加斯。 The Sphere绚烂的灯光下，黄仁勋与杨元庆并肩而立，他们之间有一种只有长期战友才懂的默契。 他们相识于20多岁的年纪，那是在个人电脑PC革命方兴未艾的年代。当年的年轻人或许都未曾预料 到，30年后，他们会在全球科技界瞩目之下再次握手，用一种名为AI工厂的新物种，去重写全球计算 基础设施的底层逻辑。 将时钟拨回至90年代中期，杨元庆在联想崭露头角之时，黄仁勋创立英伟达不久。彼时的计算世界由 Wintel(Windows+Intel)联盟主导，CPU是绝对的权力中心，显卡GPU在许多时候只是负责点亮屏幕的配 角。 联想通过极其高效的供应链管理和制造能力，一步步攀登至全球PC市场的顶峰。而英伟达则在图形处 理的荒原上艰难拓荒。两家公司的早期交集看似平淡无奇，主要是在显卡组件的采购与集成上。但正是 这种基于供应链的务实合作，为日后的深远合作埋下了伏笔。 对于联想而言，英伟达最初只是众多供应商名单中的一个名字，但随着3D游戏和多媒体需求的崛起， 联想人敏锐地意识到，单纯依靠CPU已无法满足用户对视觉体验的渴望。 联想开始在其高端台式机和后来的ThinkPad工作站系列中， ...

至于游戏玩家，英伟达或许要考虑重启旧产线（如RTX3060），移植一些新技术了。 西风 发自 凹非寺 量子位 | 公众号 QbitAI 黄仁勋最新发言： 机器人就是AI移民，能承担人类不愿意从事的工作。 现在内存卡了GPU的脖子。 在全球最大消费电子展CES 2026上，黄仁勋不仅一口气推出多款AI新品，在采访环节也是直面行业焦点。 而且老黄很忙，采访完下一站就是出席联想Tech World活动，与杨元庆 共同宣布" 联想人工智能云超级工厂 " ，英伟达最新发布的Vera Rubin将是该合作的重要组成部分 。 总之，老黄在自家发布会没说的内容，我们都整理在这了： 采访整理 CES 2026上，老黄整个核心主题都围绕着物理AI展开，包括机器人、自动驾驶等。 在机器人方面，他表示，从人口结构来看，我们已无法支撑起理想中的经济规模。因此， 需要更多的 AI新移民 ，来助力我们的生产车间 ， 承担那 些我们或许已不愿再从事的工作 。 紧接着又现身IEEE颁奖现场， 领取IEEE最高荣誉——2026年IEEE荣誉奖章 （2 026 IEEE Me dal of Honor） 。 IEEE授予黄仁勋该奖，以表彰其领先行 ...

英伟达CEO黄仁勋在接受媒体采访时指出，AI工厂的发展需要消耗大量能源，包括天然气和可再生能 源等，但"能源效率才是关键"。 （文章来源：第一财经） 他强调，英伟达每一代新产品的吞吐量提升10倍，功耗仅增加2倍，这意味着能源效率的提升。他将能 源效率与商业价值直接挂钩，数据中心的能源效率直接关系到客户的收入——在相同的电力消耗下，更 高的效率能生成更多的tokens，从而直接转化为更高的营收。他说，能源效率至关重要，英伟达持续追 求的目标是"每美元生成1个token"。 ...

Core Insights - The year 2025 marks a significant turning point for AI, transitioning from curiosity in 2024 to profound societal impact [1] - Predictions from industry leaders suggest that advancements in AI will continue to accelerate, with Sam Altman forecasting the emergence of systems capable of original insights by 2026 [1][3] - The debate around the Scaling Law continues, with some experts asserting its ongoing relevance and potential for further evolution [12][13] Group 1: Scaling Law and Computational Power - The Scaling Law has shown resilience, with computational power for training AI models growing at an exponential rate of four to five times annually over the past fifteen years [12][13] - Research indicates a clear power-law relationship between performance and computational power, suggesting that a tenfold increase in computational resources can yield approximately three times the performance gain [13][15] - The concept of "AI factories" is emerging, emphasizing the need for substantial computational resources and infrastructure to support AI advancements [27][31] Group 2: Breakthroughs in AI Capabilities - The SIMA 2 project at DeepMind demonstrates a leap from understanding to action, showcasing a general embodied intelligence capable of operating in complex 3D environments [35][39] - The ability of AI models to exhibit emergent capabilities, such as logical reasoning and complex instruction following, is linked to increased computational power [16][24] - By the end of 2025, AI's ability to complete tasks has significantly improved, with projections indicating that by 2028, AI may independently handle tasks that currently require weeks of human expertise [41] Group 3: Future Challenges and Considerations - The establishment of the Post-AGI team at DeepMind reflects the anticipation of challenges that will arise once AGI is achieved, particularly regarding the management of autonomous, self-evolving intelligent agents [43][46] - The ongoing discussion about the implications of AI's rapid advancement highlights the need for society to rethink human value in a world where intelligent systems may operate at near-zero costs [43][46] - The physical limitations of power consumption and cooling solutions are becoming critical considerations for the future of AI infrastructure [31][32]

文 | 奇点研究社，作者 | 孟雯 但和前面所述的这些成熟产品相比，它也是特殊的。根据媒体报道，字节内部人士透露，豆包在所有破 亿DAU的字节产品中，获客成本处于历史最低水平。从某种程度上，也代表豆包已经走过了"投流大 战"，走入用户自然增长的新阶段。 一天前，还有一则有关豆包的消息刷屏，在接下来的2026年春晚上，豆包将正式登台，联合火山引擎成 为央视春晚独家合作伙伴。 在国民级舞台亮相，同样的战略投入动作在字节身上从未出现过，但在互联网已不是新鲜事，如15年的 腾讯微信红包，和16年的阿里支付宝。如今，微信支付和支付宝早已经成为腾讯、阿里高市值的重要支 撑。 如果说今日头条、抖音，代表了字节最初的十年，它用算法重新为自己划分了一系列"轻模式"蛋糕—— 广告、电商。那么，从当下的种种动作来看，都可以毫不客气地说，豆包正在成为字节下一个十年的代 表。 这其中有字节对新流量入口的AII in，有对不满足于"轻资产"模式的更大野心，更有这家年轻科技企业 更激进的下一个十年的战略卡位。 豆包，正在成为"新字节"。 时隔5年，字节又一次让一款产品跑出了抖音级别的增长曲线。 上线两年半，豆包APP实现DAU破亿，成为国 ...

Core Insights - The article discusses the transition of China's AI industry from a phase of intense competition to a focus on practical applications, highlighting Lenovo's initiatives in this transformation [1][10] Group 1: AI Factory Concept - Lenovo introduced the "AI Factory" concept as a reconstruction of the computing infrastructure business model, moving beyond mere computing power to address complex industry needs [2][3] - The shift from traditional computing centers to AI factories is seen as essential for reducing costs and enhancing efficiency for small and medium enterprises [2][3] Group 2: Business Model Innovation - Lenovo aims to fill the capability gaps of its clients by providing a comprehensive suite of services, including consulting, data governance, and AI production management, thus lowering the barriers for SMEs [3][4] - The company emphasizes the importance of ecosystem collaboration to enhance operational efficiency and competitiveness [3] Group 3: Standardization Efforts - The lack of industry standards for AI training and inference efficiency is identified as a significant challenge, prompting Lenovo to collaborate with various institutions to establish integrated training and inference standards [4] - This initiative aims to eliminate information asymmetry in the computing service market, allowing clients to better understand the return on investment for their expenditures [4] Group 4: Hardware Innovations - Lenovo's new server, the WA8080a G5, is designed to address the challenges posed by the rapid evolution of GPU technology, which has seen power consumption exceed 1000 watts [5][6] - The company has adopted a modular design strategy to accommodate the fast-paced changes in GPU architecture, ensuring long-term investment protection for clients [7] Group 5: Software and Performance Optimization - Lenovo's Wanquan Heterogeneous Computing Platform 4.0 has been optimized to address emerging technology trends, particularly in handling long sequences in model training [8] - Innovations in network load balancing have been implemented to tackle bandwidth degradation issues in large-scale clusters [8] Group 6: Market Observations - Despite technological advancements, the commercial viability of AI applications remains a challenge, with only a small percentage of clients willing to pay for models [9] - The current market is compared to the mobile internet era, indicating that the ecosystem for AI applications has not yet reached a stage of widespread willingness to pay [9] Group 7: Conclusion - Lenovo's initiatives signal a significant shift in the AI computing industry towards standardized, measurable, and profitable services, marking a deep transformation in production efficiency and business models [10]

Group 1 - A total investment of nearly 10 billion yuan has been signed for 16 key projects in the new energy industry in Changzhou, which is recognized as "China's New Energy Capital" [1] - The output value of the new energy industry in Changzhou is expected to exceed 850 billion yuan in 2024, with the completeness of the power battery industry chain reaching 97%, ranking first in the country [1] - The Jintan District is a significant hub for the power battery industry in Changzhou, contributing nearly half of the city's output, with over 100 regulated enterprises in the industry chain [1] Group 2 - The Xiangfeng New Energy Project in Jintan District is a key project in Changzhou and part of the headquarters and R&D production project of Honeycomb Energy, a rising star in the lithium battery industry [1] - The project officially commenced in September 2022 and completed acceptance inspection by the end of September 2023, covering a total construction area of approximately 410,000 square meters [1] - The project is positioned as a high-standard, intelligent, zero-emission environmentally friendly AI factory, with an automation rate of 95% and equipped with world-class battery automation production equipment [1] Group 3 - Changzhou is focusing on the new energy industry as the core of its modern energy system construction during the 14th Five-Year Plan period, increasing efforts in technological innovation [2] - The city aims to attract projects worth 100 billion yuan, cultivate enterprises worth 1 trillion yuan, and build an industry worth 10 trillion yuan [2] - By constructing a green lifestyle centered on new energy utilization, Changzhou is injecting strong momentum into the high-quality development of its modern energy system [2]

Group 1 - The core discussion revolves around the future of money and work in the context of advanced AI and robotics, with Elon Musk suggesting that money may become irrelevant as AI leads to unprecedented material abundance [1][2] - Musk envisions a future where work becomes optional and driven by passion rather than necessity, drawing parallels to hobbies like gardening [2][4] - Jensen Huang offers a more cautious perspective, asserting that while AI will change the nature of work, it will not eliminate the need for work altogether, and people may become busier as AI enhances productivity [3][4] Group 2 - The partnership between HUMAIN and Musk's xAI to build multiple super data centers in Saudi Arabia, including a massive 500 MW facility, highlights the region's ambition to become a global AI infrastructure hub [4][5] - Saudi Arabia's low energy costs, vast land, and capital availability position it as a strategic player in the AI landscape, aiming for a total capacity of 1.9 GW by 2030 [5][6] - Huang describes these super data centers as "AI factories," emphasizing their role in producing AI models and content rather than merely storing data [6][7] Group 3 - Huang identifies three key factors driving the AI boom: the need for processing vast amounts of data, the shift from recommendation algorithms to generative AI, and the rise of autonomous intelligent agents [8][9] - He argues that the current demand for AI computing power is based on real needs and technological evolution, distinguishing it from past tech bubbles [10] - The collaboration between Chinese companies and Saudi Arabia in AI infrastructure development reflects a growing trend of international partnerships in the tech sector [11][12] Group 4 - Geopolitical factors pose challenges to Sino-Saudi AI cooperation, particularly concerning U.S. restrictions on high-end AI technology exports [12][13] - The technological gap in high-performance computing may necessitate a focus on application-level collaborations rather than direct competition with U.S. firms [13][14] - Saudi Arabia aims to balance its partnerships with both U.S. and Chinese companies to maximize its technological and economic benefits [14]

Core Viewpoint - The appointment of Dr. Huang Chenhong as Chairman and CEO of SuperX AI Technology Limited marks a strategic milestone for the company as it aims to become a global leader in AI infrastructure solutions, leveraging his extensive experience in multinational technology firms [1][2][10]. Group 1: Leadership Transition - Dr. Huang's appointment is not just a routine executive change but a significant step for SuperX in its pursuit of leadership in the AI infrastructure sector [1]. - His extensive background includes over 30 years in enterprise software, IT hardware, data center infrastructure, and network communications, making him a well-suited leader for SuperX's ambitious goals [3][5]. - The leadership transition aligns with SuperX's mission to build an integrated, end-to-end AI infrastructure solution, which requires a leader with a deep understanding of complex operations and global market strategies [2][6]. Group 2: Dr. Huang's Qualifications - Dr. Huang holds a Ph.D. in Electrical Engineering from Texas A&M University and has held significant positions at SAP, Dell, and Schneider Electric, showcasing his capability in managing large-scale technology operations [3][4]. - His experience at Dell, particularly in AI server and data center modernization, positions him to enhance SuperX's product development and competitive edge in the AI infrastructure market [4][5]. - Dr. Huang's expertise in data center infrastructure and power management is crucial for SuperX, especially in optimizing their proprietary liquid cooling solutions for high-performance AI [6][7]. Group 3: Strategic Vision and Goals - Under Dr. Huang's leadership, SuperX aims to accelerate the global deployment of its AI infrastructure, focusing on operational execution and establishing itself as a leader in the AI infrastructure space [10][11]. - The company has outlined four key priorities for fiscal year 2026, including integrating recent acquisitions, advancing AI factory deployments, strengthening partnerships, and expanding global service capabilities [9][11]. - The recent appointments of other board members, such as Frank Han and Roy Rong, further enhance SuperX's governance and strategic execution capabilities, aligning with its vision for growth and market leadership [8][9].

Core Insights - South Korean President Lee Jae-myung emphasizes the central role of artificial intelligence (AI) in the government's economic vision, aiming for comprehensive integration of AI across core industries, public services, and defense [1][2] - Major South Korean conglomerates, including Samsung and Hyundai, are collaborating with NVIDIA to establish AI factories, significantly boosting the domestic AI infrastructure [1][4] - The KOSPI index has seen substantial gains, with a monthly increase of over 20% in October, driven by the positive sentiment surrounding AI investments [1] Government Initiatives - The South Korean government plans to increase AI investment to 10.1 trillion KRW (approximately 7 billion USD), more than tripling the current investment level, as part of a broader budget proposal of 728 trillion KRW [2] - AI is identified as a foundational technology for key sectors such as humanoid robotics, autonomous vehicles, and semiconductors, with a focus on integrating real-time data with the manufacturing base [2] - The government aims to enhance the national computing power system and develop a larger talent pool in AI technology [2] Defense Spending - The defense budget is set to increase by 8.2% to 66.3 trillion KRW, with a focus on upgrading conventional weapons and introducing AI-driven military systems [3] - The military spending will account for 2.4% of the GDP, reflecting a commitment to national pride in self-defense capabilities [3] Corporate Collaborations - NVIDIA is partnering with Samsung, SK Group, and Hyundai to build AI factories, with plans to supply over 260,000 high-performance AI GPUs for these projects [4][5] - Samsung's AI factory will feature more than 50,000 NVIDIA GPUs, and discussions are ongoing regarding the supply of next-generation high-bandwidth memory chips [5] - The collaboration with SK Group aims to enhance semiconductor R&D and cloud infrastructure, with the first phase of the AI factory expected to be completed by the end of 2027 [5] - The partnership with Hyundai focuses on developing AI capabilities for autonomous vehicles and smart factories, deploying 50,000 NVIDIA GPUs for integrated AI model training and deployment [6]