2026年国际消费电子展(CES)将于1月6日在拉斯维加斯正式拉开帷幕，届时英伟达掌门人黄仁勋、AMD 首席执行官苏姿丰将发表重要演讲。阿里巴巴、联想、三星电子、LG等亚洲科技力量将集中亮相 CES，事关英特尔代工业务命运的首款"18A"芯片届时也将浮出水面。CES 2026的主题依旧围着AI而展 开，但变化的是，AI将转入场景落地阶段。外界一致认为，AI硬件应用有望成为本次CES的重中之重。 AI眼镜方面，CES 2026的AI眼镜厂商超50余家，雷鸟创新、Rokid、影目、VITURE、XREAL等头部厂 商悉数出席，此外，闪极、BleeqUp、Halliday、微光科技等新锐品牌也开始展露头角。阿里巴巴将携旗 下首款自研 AI 眼镜——夸克 AI 眼镜 S1 ，亮相 CES 2026。 AI 机器人方面，特斯拉第三代人形机器人Optimus-3或许将在CES迎来首秀。宇树科技据称将在CES 2026 带来人形机器人的最新逼真交互演示。另外，灵犀智能将首次亮相CES 2026，展示其第一款重磅 产品——AiMOON星座AI守护精灵。智元机器人亦有望在CES 2026上展出灵犀X2(AGIBOT X2)、远 ...

以下文章来源于硬AI ，作者专注科技产研的 硬AI . AI时代，快人一步~ 来源 | 硬AI 芯片巨头的策略在CES 2026将出现显著分化，英伟达的重心明显不在传统的消费级显卡市场。 具身智能战场 从中美韩"三国杀"看执行力进阶 2026年被视为人形机器人从Demo走向工位的关键年份，核心看点在于"降本增效"与"实战验证"。 端侧 AI 与 XR 安卓阵营的反击 与中国厂商的"轻量化"突围 XR市场正在经历Apple Vision Pro发布后的"消化期"，CES 2026将是安卓阵营反击的起点。 汽车产业 从"软件定义"跃迁至"AI 定义" 芯片军备竞赛升级 汽车行业正经历技术架构的深层变革，智驾芯片的竞争格局日益清晰。 英伟达（NVIDIA） ： 市场预期的RTX 50 Super系列显卡（包括RTX 5080 SUPER等）或因GDDR7显存价格高企和供应短缺而 推迟 发布 。CEO黄仁勋的演讲将聚焦"Physical AI"， 推动AI算力向机器人与工业场景延伸 。英伟达可能转而专注于高利润产品，如96GB 显存的RTX PRO 6000 Blackwell。 AMD ： 采取稳健升级策略。 桌面 ...

CES 2026凸显芯片巨头战略分野：英伟达重心转向工业AI与机器人，而AMD与英特尔则坚守并升级传统PC市场。展会 焦点已从消费电子转向人形机器人、AI定义汽车及卷轴屏等颠覆性硬件的产业化竞速，中、美、韩厂商在各赛道展开激烈 角逐。 硬·AI 作者 |叶慧雯 编辑 | 硬 AI 2026年消费电子展（CES）将于1月4日至9日在美国拉斯维加斯举行，CES一向是科技企业发表年度新品 的重要舞台，展品涵盖即将上市的产品，以及仍处于概念阶段、未必投产的装置。 芯片巨头的策略在CES 2026将出现显著分化，英伟达的重心明显不在传统的消费级显卡市场。 中国阵营（成本与量产）： 宇树和智元继续主导成本控制。宇树或展示G1量产版在工厂流水线的 操作；智元将展示X2、A2等全系列产品及灵巧手等核心部件。 美国阵营（技术标杆）： 波士顿动力的全电动Atlas将进行首次公开演示，直接对标特斯拉 Optimus，展示商业化潜力。 韩国阵营（产业链抱团）： 韩国将展示"K-Humanoid"联盟，由三星投资的Rainbow Robotics领 衔。展品包括轮式底盘工业级人形机器人HMND 01 Alpha（身高220厘米，载荷1 ...

Core Insights - Google Cloud has officially launched its Google Cloud G4 VMs, powered by NVIDIA's RTX PRO 6000 Blackwell GPUs, aimed at enhancing AI applications in industrial and enterprise settings [1][2][3] - The G4 VMs offer up to 9 times the throughput compared to the previous G2 platform, significantly improving performance for various AI workloads [2][4] - The collaboration between Google and NVIDIA establishes a comprehensive cloud platform that supports both AI training and physical AI workloads, catering to a broader range of enterprise needs [4][5] Product Features - The G4 VMs utilize NVIDIA's RTX PRO 6000 Blackwell GPUs, which combine advanced Tensor Cores and RT Cores for enhanced AI performance and real-time rendering capabilities [3][6] - The integration of Google Kubernetes Engine and Vertex AI simplifies the deployment of containerized applications and machine learning operations [3][4] - The G4 VMs are designed to support a wide range of workloads, including multimodal AI inference, digital twins, and complex visual computing [5][6] Market Impact - The introduction of G4 VMs is expected to drive significant growth for both Google and NVIDIA, as it addresses the increasing demand for AI capabilities in various industries [7][8] - NVIDIA's stock is projected to continue rising, with analysts predicting a potential market capitalization exceeding $5 trillion within a year [7][8] - The AI infrastructure investment wave is anticipated to reach between $2 trillion to $3 trillion, driven by the demand for AI computing resources [9]

Core Insights - Google Cloud has officially launched its Google Cloud G4 VMs, powered by NVIDIA's RTX PRO 6000 Blackwell GPUs, aimed at enhancing AI applications across various industries [1][2][3] - The G4 VMs offer up to 9 times the throughput compared to the previous G2 platform, significantly improving performance for multimodal AI workloads and complex simulations [2][5] - NVIDIA's Omniverse and Isaac Sim platforms are now available on Google Cloud Marketplace, providing essential tools for industries like manufacturing and logistics [2][6] Product Features - The G4 VMs utilize NVIDIA's RTX PRO 6000 Blackwell GPUs, which feature fifth-generation Tensor Cores and fourth-generation RT Cores, enhancing AI performance and real-time ray tracing capabilities [3][5] - The integration of Google Kubernetes Engine and Vertex AI simplifies the deployment of containerized applications and machine learning operations for physical AI workloads [3][4] - G4 VMs are designed to cater to a broader range of enterprise workloads, particularly those requiring low-latency AI inference and digital twin simulations [5][6] Market Impact - The introduction of G4 VMs is expected to drive significant growth for both Google and NVIDIA, as they establish a comprehensive cloud computing platform for AI training and inference [3][7] - NVIDIA's strong position in the AI computing market is reinforced by its partnerships and investments, including a substantial deal with OpenAI [7][8] - Analysts predict that NVIDIA's stock will continue to rise, with target prices being adjusted upwards, indicating a bullish outlook for the AI infrastructure market [7][8] Industry Trends - The AI computing sector is experiencing a surge in investment, with estimates suggesting a potential market size of $2 trillion to $3 trillion driven by unprecedented demand for AI infrastructure [8][9] - The recent price increases in high-performance storage products and strong earnings from key players like TSMC further support the bullish narrative for AI-related hardware and infrastructure [9]

Core Insights - Google Cloud has officially launched its Google Cloud G4 VMs, powered by NVIDIA's RTX PRO 6000 Blackwell GPUs, aimed at enhancing AI applications in industrial and enterprise settings [1][2][3] - The G4 VMs offer up to 9 times the throughput compared to the previous G2 platform, significantly improving performance for various AI workloads [2][5] - NVIDIA's Omniverse and Isaac Sim platforms are now available on Google Cloud Marketplace, providing essential tools for industries like manufacturing and logistics [2][6] Product Features - The G4 VMs utilize NVIDIA's RTX PRO 6000 Blackwell GPUs, which feature advanced Tensor Cores and RT Cores for enhanced AI performance and real-time ray tracing capabilities [3][5] - The integration of Google Kubernetes Engine and Vertex AI simplifies the deployment of AI workloads, making it easier for users to manage machine learning operations [3][5] - The G4 VMs are designed to cater to a wide range of enterprise AI workloads, including low-latency inference and digital twin applications [5][6] Market Impact - The introduction of G4 VMs is expected to lower the entry barrier for enterprises looking to adopt AI technologies, thus expanding the market for AI inference workloads [5][6] - NVIDIA is positioned as a key beneficiary of the ongoing AI spending wave, with analysts projecting significant stock price increases and market capitalization growth [7][10] - The global AI infrastructure investment is anticipated to reach between $2 trillion and $3 trillion, driven by unprecedented demand for AI computing power [10]

Core Insights - Nvidia's latest GPU architecture, Blackwell, features the largest chip, GB202, with a die size of 750 mm² and 92.2 billion transistors, designed for high performance in graphics processing [1][62] - The RTX PRO 6000 Blackwell configuration is the most powerful in Nvidia's lineup, comparable to the RTX 5090 but with more stream multiprocessors (SMs) enabled [1][2] Architecture and Performance - The GB202 chip has 192 SMs, which are the fundamental building blocks of Nvidia GPUs, and utilizes a large memory subsystem to enhance performance [1][4] - Blackwell's SM to GPC ratio is 1:16, allowing for cost-effective scaling of SMs without increasing GPC-level hardware [5] - Compared to AMD's RDNA4 architecture, which has a 1:8 SE:WGP ratio, Blackwell's design allows for higher clock speeds and potentially greater throughput [6][18] Instruction and Execution - Blackwell uses fixed-length 128-bit instructions and a two-level instruction cache, improving instruction bandwidth and performance [7][10] - The architecture allows for overlapping different types of workloads in the same queue, enhancing efficiency in shader array utilization [8][23] Memory Subsystem - Blackwell features a 128 KB memory block divided into L1 cache and shared memory, maintaining low latency and high throughput [25][35] - The L2 cache latency is slightly higher than previous generations, but the overall memory bandwidth remains superior to AMD's offerings [49][53] Competitive Landscape - Nvidia's RTX PRO 6000 Blackwell outperforms AMD's RX 9070 in various benchmarks, particularly in memory bandwidth and computational performance [58][61] - The competition in the GPU market is intensifying, with Intel's upcoming Battlemage and AMD's RDNA4 targeting mid-range markets, while Nvidia continues to dominate the high-end segment [61][64]