作者丨 新智元 编辑丨 新智元 点击下方 卡片 ，关注" 具身智能之心 "公众号 >> 点击进入→ 具身 智能之心 技术交流群 更多干货，欢迎加入国内首个具身智能全栈学习社区 ： 具身智能之心知识星球 (戳我) ， 这里包含所有 你想要的。 【导读】 2025年，全球具身智能赛道爆火，VLA模型成为了绝对的C位。从美国RT-2的开创性突 破，到中国最新FiS-VLA「快慢双系统」，VLA正以光速硬核进化。 2025年，具身智能可真是太火了。 而提到具身智能，不得不提——视觉语言动作模型（Vision-Language-Action，VLA）。 作为具身智能的核心驱动力，VLA正席卷全球，成为研究人员们的「新宠」。 论文链接: https://arxiv.org/pdf/2506.01953 从产业界到学术界，全球的主流公司与研究机构，都在加速向这一方向靠拢，达成了罕见的共识。 在硅谷，诸如谷歌DeepMind、Figure AI、Skild AI、Physical Intelligence等行业领军者，早已 开始发力押注VLA的未来。 几周前，谷歌曾发布了首个离线VLA模型，让机器人不用联网，即可精准操控完成 ...

英伟达Q1收入为440亿美元（上篇一致预期434亿美金），同比增长69%，Non-GAAP净利润187.75亿美 元，同比+26%、环比-15%，主要由于H20带来的45亿美元费用影响。 数据中心收入390亿，同比增长73%。AR工作负载已 强烈转变为推理 ，而人工智能工厂扩建正在推动 可观的收入。我们对客户的承诺是坚定的。4月9日，美国政府发布了新的H20出口管制措施。 Q2 总收入预计为450亿+/-2%。（市场近期调高一致预期465亿美金，此前450亿）。请注意，我们的展 望反映了H20收入损失约80亿美元。 数据来源：彭博 数据中心业务 ：25Q1实现收入391.12亿美元，同比+73%、环比+10%，略低于市场预期（392亿美 金），25Q1中H20销售额为46亿美元，受出口限制，无法发货的H20为25亿美元。 游戏和AI PC业务 ：25Q1实现收入37.63亿美元，同比+42%、环比+48%，超市场预期（28.4亿美元）。 3）CSP开始对GB 300系统进行Ultra采样，我们预计将向Commerce，Commerce发货。GB300的下拉式 设计将允许CSPS无缝过渡。他们的系统和制造用于G ...

Core Viewpoint - Embodied intelligence is a key focus in the AI field, particularly in humanoid robots, raising questions about the best path to achieve true intelligence and the current challenges in data, computing power, and model architecture [2][5][36]. Group 1: Development Stages of Embodied Intelligence - The industry anticipates 2025 as a potential "year of embodied intelligence," with significant competition in multimodal and embodied intelligence sectors [5]. - NVIDIA's CEO Jensen Huang announced the arrival of the "general robot era," outlining four stages of AI development: Perception AI, Generative AI, Agentic AI, and Physical AI [5][36]. - Experts believe that while progress has been made, the journey towards true general intelligence is still ongoing, with many technical and practical challenges remaining [36][38]. Group 2: Transition from Autonomous Driving to Embodied Intelligence - Many researchers from the autonomous driving sector are transitioning to embodied intelligence due to the overlapping technologies and skills required [17][22]. - Autonomous driving is viewed as a specific application of robotics, focusing on perception, planning, and control, but lacks the interactive capabilities needed for general robots [17][19]. - The integration of expertise from autonomous driving is seen as a bridge to advance embodied intelligence, enhancing technology fusion and development [18][22]. Group 3: Key Challenges in Embodied Intelligence - Current robots often lack essential capabilities, such as tactile perception, which limits their ability to maintain balance and perform complex tasks [38][39]. - The operational capabilities of many humanoid robots are still in the demonstration phase, lacking the ability to perform tasks in real-world contexts [38][39]. - The complexity of high-dimensional systems poses significant challenges for algorithm robustness, especially as more sensory channels are integrated [39]. Group 4: Future Applications and Market Focus - The focus for developers should be on specific application scenarios rather than pursuing general capabilities, with potential areas including home care and household services [48]. - Industrial applications are highlighted as promising due to their scalability and the potential for replicable solutions once initial systems are validated [48]. - The gap between laboratory performance and real-world application remains significant, necessitating a focus on improving system accuracy in specific contexts [46][47].

具身智能，作为近年来人工智能领域的热点之一，成为产业界和学术界重点关注的方向。特别是在人形机器人这个载体上，它所承载的感知、运 动、决策等能力，让具身智能从概念逐渐走向落地。但与此同时，也有不少值得深入探讨的问题浮出水面：为什么具身智能的发展似乎格外偏 爱"人形"？是否只有模仿人类形态，才是实现智能的最佳路径？在面对数据、算力、模型架构等现实挑战时，我们究竟处于怎样的阶段？距离真 正的通用机器人，还有多少"里程"要走？ 基于此，CSDN《万有引力》栏目特别策划了一期以"十问具身智能：我们离通用机器人还有多远？"为主题的深度对话，邀请了北京邮电大学人 工智能学院副教授陈光@爱可可-爱生活、深圳市人工智能与机器人研究院副研究员夏轩、Roboraction.AI 首席执行官黄浴，在栏目主理人 CSDN &《新程序员》执行总编唐小引主持下，三位专家将从技术演进、研究现状、产业应用等多个角度切入，带大家一同拆解具身智能面临的"关键问 题"，看清这条通往未来机器人的发展路径。 夏轩：在专业背景方面，我早期的研究主要集中于计算机视觉领域（CV），涵盖无人机图像处理、工业图像处理以及生成模型等方向。在扩散模 型兴起之前，我也 ...

点击招商研究小程序查看PDF报告原文 事件： 英伟达CEO黄仁勋5月19日在COMPUTEX 2025大会发表主题演讲，介绍公司在AI基础设施和"物理AI"领域最新进展，以及Blackwell等硬件平台 与机器人等技术前沿应用。综合演讲及材料信息，总结要点如下： 评论： 1、构建智能基础设施迈向物理AI，5G/6G以及量子计算持续推进。 英伟达表示未来智能基础设施建立在电力和互联网基础上，AI从感知和推理向自主决策演进，未来将实现物理AI，AI驱动的机器人执行现实世界任 务，英伟达发展历程从芯片公司到AI基础设施公司，ShonaAriel、Kulitho、DynamoQDF等函数库，科学领域、物理领域加速应用，持续推进AI向 5G/6G、量子计算领域迈进。 2、GB300推理性能提升1.5倍，NV Link Fusion提供半定制芯片平台。 1）GB300： 英伟达25Q3将推出GB300，在保持相同架构、物理规格和机电设计基础上，相较GB200实现推理性能提升1.5倍，HBM增加1.5倍， 网络性能翻倍，训练性能持平；GB300中央区域采用全液冷设计，外部接口保持兼容，单节点达40petaflops，采用台 ...

今天，老黄在Computex 2025大会上，发表了一场长达两小时的主题演讲。 一开始老黄回顾了Nvidia 的发展历程，从专注于GPU，到2006年推出CUDA，再到AI基础设施巨 头，其实这场演讲中提到的很多产品之前就推出了，只是在这场演讲中又提到了一些细节。 在这次的演讲中，最吸引我的还是NVLink Fusion。这篇文章就分析一下这个技术。 进入正文之前，先扯点别的。 老黄确实非常会演讲，当听到上面这段话的时候，真心佩服老黄。把英伟达带到了这样一个高度。 可以说如果没有英伟达，AI的发展进程不会有这么快。 但也不知为什么，耳边还是会经常响起Linus的那句：Fuck Nvidia. GB300计划在Q3推出，该芯片推理性能提升 1.5 倍、HBM内存提升 1.5 倍、网络带宽提升 2 倍， 并与上一代保持物理兼容性，实现100%液冷。 CES上提及的Project DIGITS的个人AI计算机DGX Spark已全面投产，老黄表示每个人都可以在圣 诞节拥有一台。 RTX Pro 企业 AI 服务器，支持传统x86、Hypervisor、Windows 等 IT 工作负载----笔者对这个产 品一直都 ...

黄仁勋最新演讲！ 5月19日，英伟达CEO黄仁勋在台北国际电脑展发布了一系列英伟达公司在软硬件方面的更新，并介绍了 他在未来全球AI（人工智能）领域的宏大畅想。 AI无处不在 英伟达芯片将是基础设施 黄仁勋在演讲开始就谈起了他对AI未来的伟大期许：未来AI将会像互联网和电力一样，成为我们生活中不 可或缺的必要组成部分。 "我知道，现在当我们讨论AI基建时，你可能会感觉到好像没什么必要。但我向你保证，十年后你会意识 到，我们需要无处不在的人工智能！"黄仁勋在演讲中谈道："我们会将AI集成到所有地方：每个地区、每 个行业、每个工厂、每个公司，全都需要AI！" 他畅想AI前景时表示："我们正处于繁荣未来的边缘，芯片产业的价值已达3000亿美元，而数据中心的机 遇正在转变为近万亿美元的市场，这一切受到人工智能工厂和基础设施的推动。" 他同时解释了英伟达在这场AI基建中所扮演的角色："到那时，AI就像是互联网和电力一样，它将需要工 厂！而这正是我们现在正在制造的……英伟达不再仅仅是一家科技公司，它是一家重要的基础设施公 司。" 高度评价DeepSeek 在演讲中，黄仁勋再次对DeepSeek表示赞赏。他表示，DeepS ...

Group 1 - Nvidia's CEO Jensen Huang emphasized the impracticality of training robots in the physical world, advocating for training in a virtual environment that adheres to physical laws [1] - Nvidia is collaborating with DeepMind and Disney Research to develop a cutting-edge physics engine called Newton, which will be open-sourced in July [1] - The Newton engine supports GPU acceleration and features high differentiability and ultra-real-time operation capabilities, enabling effective learning through experience [1] Group 2 - Nvidia is advancing robotic systems in the automotive industry using the Isaac Groot platform, powered by a new processor named Jetson Thor, designed for a wide range of robotic applications [3] - The Isaac operating system manages neural network processing, sensor processing, and data pipelines, enhancing system capabilities with pre-trained models developed by specialized teams [3] - Nvidia's end-to-end autonomous vehicle technology stack is a comprehensive solution developed entirely in-house, facilitating collaboration with various partners to promote the commercialization of autonomous vehicles [3] Group 3 - Nvidia is applying its AI models to autonomous vehicles, launching a fleet in collaboration with Mercedes globally, aiming for implementation of end-to-end autonomous driving technology this year [5]

在今日举行的技术发布会上，英伟达(NVDA.O)创始人兼CEO黄仁勋发表一系列重磅声明，全面描绘了 英伟达在人工智能时代的新战略定位。他表示，英伟达已不再只是传统意义上的科技公司，而是"AI基 础设施公司"，将在全球构建面向未来的计算、机器人和智能系统生态，"AI基础设施市场将以数万亿美 元计量，tokens产能未来将以小时为单位衡量。" 推出多款新品，构建全面AI生态 黄仁勋宣布推出新一代AI个人计算设备DGX Spark与DGX Station工作站，并将Blackwell RTX Pro6000主 板推向量产，配合NVLink CX8互联技术，打造支持800Gbps通信带宽的高速GPU集群系统。同时，新 一代超算架构也同步曝光，其带宽高达130TB，搭载72颗处理器或144颗GPU，构建起全球最大规模的 人工智能算力平台之一。 AI机器人"全民化"：移动设备将变为智能体 黄仁勋强调，未来所有移动设备都将成为机器人，这将引发工业革命。"我们正在与汽车行业并行推进 Isaac Groot机器人系统，该系统由Jetson Thor驱动，可适配自动驾驶与人机协作平台。" 英伟达已与梅赛德斯合作，在全球部署搭载端 ...

Summary of NVIDIA 2025 Conference Call Company Overview - **Company**: NVIDIA (NVDA) - **Event**: 2025 Conference held on May 18, 2025 Key Industry Insights - NVIDIA is positioned at the center of the computer ecosystem, emphasizing its role in creating new markets and growth opportunities [2][3] - The company has transitioned from a chip manufacturer to an essential infrastructure company, particularly in AI [12][13] - The concept of AI infrastructure is compared to historical infrastructures like electricity and the Internet, indicating its future significance [14][16] Core Product Developments - Introduction of new products aimed at revolutionizing computing, particularly in AI and accelerated computing [22][24] - The launch of the **GeForce RTX 50 series**, which achieved the fastest launch in NVIDIA's history, highlighting the growth of PC gaming [28] - Development of **Grace Blackwell**, a new system designed for inference time scaling, which is now in full production [60][61] Technological Innovations - NVIDIA's focus on **accelerated computing** and the importance of libraries, particularly CUDA, in driving innovation [22][30] - Introduction of **NVLink Fusion**, allowing for semi-custom AI infrastructure, enabling integration with various CPUs and ASICs [87][90] - The **DGX Spark** and **DGX Station** are designed for AI-native developers, providing powerful computing capabilities for research and development [97][100][103] Market Opportunities - Emphasis on the telecommunications industry transitioning to software-defined networks, with partnerships for AI integration in 5G and 6G technologies [40][41] - The potential for AI to transform various industries, including telecommunications, genomics, and medical imaging [34][40] Future Vision - NVIDIA envisions a future where AI is integrated into every aspect of infrastructure, similar to how electricity and the Internet became essential [16][17] - The concept of **agentic AI**, which can reason and act, is highlighted as a significant advancement in AI capabilities [50][52] - The company aims to reinvent enterprise IT by integrating AI capabilities into traditional computing environments [108][112] Financial and Market Impact - The AI infrastructure market is projected to be a trillion-dollar opportunity, with NVIDIA's role as a key player in this transformation [21][22] - The company is building AI factories, indicating a shift from traditional data centers to more advanced computing environments [73][74] Partnerships and Collaborations - Collaboration with major companies like TSMC, Foxconn, and various telecommunications firms to enhance AI infrastructure and capabilities [39][42][95] - NVIDIA's ecosystem includes partnerships with companies like Dell, HPI, and ASUS for product development and distribution [98][99] Conclusion - NVIDIA is at the forefront of AI and computing innovation, with a clear roadmap for the future that emphasizes the integration of AI into all aspects of technology and infrastructure [12][13][21]