--- 格式版本: 2 标题: "Nvidia has big plans to expand its 'physical AI' business" 原文链接: "https://www.linkedin.com/news/story/nvidia-has-big-plans-to-expand-its-physical-ai-business-7548356/" 发布日期: "2026-08-30" 发布时间校准状态: "found" 发布时间需复核: "否" 发布时间来源: "rule:scrape:strict_html_body" 发布时间证据: "div class=base-main-card__info self-center ml-1 flex-1 relative break-words papabear:min-w-0 mamabear:min-w-0 babybear:w-full: Jason Seligson 发布日期: 2026年8月30日" 发布时间校准原因: "规则确认唯一严格发布时间，来源 scrape:strict_html_body" 发布时间校准置信度: "high" 发布时间候选数量: 11 发布时间严格候选数量: 3 发布时间原页读取状态: "source template page reused from URL open" 发布时间未找到原因: "" 发布时间校准时间: "2026-09-01T01:55:23+08:00" 发布时间仲裁状态: "skipped" 发布时间仲裁尝试次数: 0 发布时间仲裁耗时毫秒: 0 发现时间: "2026-09-01T01:54:37+08:00" 入库时间: "2026-08-31T17:55:23.536Z" 来源平台: "LinkedIn 内容搜索" 搜索渠道: "source_template" 搜索词: "site:linkedin.com NVIDIA" 匹配关键词: - "AI" 相关厂家: - "NVIDIA" 相关专家: [] 内容类型: "网页" 抓取工具: "Free Fetch + Defuddle" 清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取" 原始附件: [] AI优质: "否" AI打分: 23 AI分档: "非优质" AI质检状态: "不通过" AI打分理由: "内容聚焦NVIDIA物理AI业务（机器人、汽车、无人机），与超节点/AI Rack/机柜级基础设施无关，无架构、互连、供电、散热等技术细节，仅提及营收预期，非项目关注方向。" AI质检模型: "tx-deepseek-v4-flash" AI质检时间: "2026-09-01T01:55:42+08:00" AI主题相关性: 2 AI来源权威性: 5 AI新颖性: 8 AI技术细节: 0 AI商业部署信号: 3 AI完整性: 5 AI摘要: "Nvidia正扩展“物理人工智能”业务，将芯片技术用于机器人、汽车和无人机等实体设备；该业务目前年收入约100亿美元，CEO黄仁勋预计未来十年将增长10倍。" AI摘要模型: "ali-deepseek-v4-flash" AI摘要时间: "2026-08-31T23:36:57.695Z" 采集批次: "2026年8月31日23点05分08秒" 采集批次ID: "20260831-230508-125" 去重键: "https://www.linkedin.com/news/story/nvidia-has-big-plans-to-expand-its-physical-ai-business-7548356" --- ## Nvidia has big plans to expand its 'physical AI' business ### Jason Seligson Nvidia is setting its sights on the next frontier of artificial intelligence. According to [The Wall Street Journal](https://www.wsj.com/tech/ai/nvidia-wants-to-run-the-worlds-robots-china-is-an-eager-customer-bdf46169), the chip giant is expanding its “physical AI” business — pairing its technology with real-world machines like robots, cars, and drones. Nvidia founder and CEO [Jensen Huang](https://www.linkedin.com/in/jenhsunhuang/) expects the physical AI division, which currently generates $10 billion in annual revenue, to grow “10-fold” over the next decade, per the outlet. Analysts see China as crucial to Nvidia’s future in physical AI. While the U.S. has banned Nvidia's most-advanced AI chips from going to China, trade "remains open for robot and car chips." --- ## 领英热点 --- 1. [NVIDIA](https://www.linkedin.com/company/nvidia?trk=news_storyline_guest-text) wants to run the world’s robots. Fine. WSJ’s Raffaele Huang explains how. But the most interesting part of this picture is not domination. It is interdependence. American chips and open-weight models. Chinese manufacturing scale and real-world use cases. Developers and robot makers across borders turning the technology into something that can actually operate in the physical world. This is open innovation at its most concrete. Not openness as ideology. Open innovation as the only possible path to making AI genuinely useful to humankind, and therefore sustainably profitable. No company, country or model can build physical AI alone. The intelligence may start inside a lab, but its value emerges through an ecosystem. One question, though: where does Figure fit into this wider picture? [Brett Adcock](https://www.linkedin.com/in/brettadcock?trk=news_storyline_guest-text) from everything I have seen, Figure is building some of the most disruptive physical AI technology so far. I would be genuinely curious to understand its place in this emerging ecosystem. Perhaps the real race is not about who controls the world’s robots. It is about who builds the ecosystem through which they can learn, work and create value for humans. [#ai](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fai&trk=news_storyline_guest-text) [#physicalai](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fphysicalai&trk=news_storyline_guest-text) [#robotics](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Frobotics&trk=news_storyline_guest-text) [#openinnovation](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fopeninnovation&trk=news_storyline_guest-text) 2. Nvidia already powers the AI behind our screens. Now, it wants to power the AI in the real world. Robots. Self-driving cars. Drones. Factories. Nvidia is betting big on what it calls “Physical AI”, AI that doesn't just generate answers but can understand and interact with the physical world. The company says this business already generates around $10 billion annually and Jensen Huang expects it to grow 10x over the next decade. The next AI revolution may not happen on your screen. It might be walking, driving, and working right beside you. [#AI](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fai&trk=news_storyline_guest-text) [#Nvidia](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fnvidia&trk=news_storyline_guest-text) [#Robotics](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Frobotics&trk=news_storyline_guest-text) [#PhysicalAI](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fphysicalai&trk=news_storyline_guest-text) [#ArtificialIntelligence](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fartificialintelligence&trk=news_storyline_guest-text) [#Technology](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Ftechnology&trk=news_storyline_guest-text) [#FutureOfAI](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Ffutureofai&trk=news_storyline_guest-text) 3. Nvidia Wants to Become the Reference Platform for “Physical AI” The most interesting development this Sunday comes from The Wall Street Journal. Nvidia’s physical AI business, which includes robots, vehicles, and drones, is already generating approximately $10 billion in annual revenue. Jensen Huang believes this business could grow tenfold over the next decade. The strategy closely mirrors the approach Nvidia used in data centers: the company is not simply trying to sell chips. It is building a vertically integrated platform combining processors, software tools, and specialized models, including GR00T and Cosmos, with the ambition of becoming the common technology layer powering the next generation of robots. At the same time, China is emerging as one of Nvidia’s key expansion markets in this field. 4. [Vishal Mehta](https://in.linkedin.com/in/vishal-mehta-686583114?trk=news_storyline_guest_feed-reaction-header) 评论了 [NVIDIA](https://www.linkedin.com/company/nvidia?trk=news_storyline_guest-text) wants to run the world’s robots. Fine. WSJ’s Raffaele Huang explains how. But the most interesting part of this picture is not domination. It is interdependence. American chips and open-weight models. Chinese manufacturing scale and real-world use cases. Developers and robot makers across borders turning the technology into something that can actually operate in the physical world. This is open innovation at its most concrete. Not openness as ideology. Open innovation as the only possible path to making AI genuinely useful to humankind, and therefore sustainably profitable. No company, country or model can build physical AI alone. The intelligence may start inside a lab, but its value emerges through an ecosystem. One question, though: where does Figure fit into this wider picture? [Brett Adcock](https://www.linkedin.com/in/brettadcock?trk=news_storyline_guest-text) from everything I have seen, Figure is building some of the most disruptive physical AI technology so far. I would be genuinely curious to understand its place in this emerging ecosystem. Perhaps the real race is not about who controls the world’s robots. It is about who builds the ecosystem through which they can learn, work and create value for humans. [#ai](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fai&trk=news_storyline_guest-text) [#physicalai](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fphysicalai&trk=news_storyline_guest-text) [#robotics](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Frobotics&trk=news_storyline_guest-text) [#openinnovation](https://www.linkedin.com/login?session_redirect=https%3A%2F%2Fwww.linkedin.com%2Ffeed%2Fhashtag%2Fopeninnovation&trk=news_storyline_guest-text)