Tag
Nvidia CEO Jensen Huang's GTC Berlin keynote demo of Pollen Robotics' adorable Microduck robots went viral, pushing preorders past 20,000 units and positioning them as a potential hot Christmas toy. The open-source robots, trained with Nvidia's simulation stack, come from Pollen Robotics, which was acquired by Hugging Face before Nvidia bought Hugging Face in September.
Kazakhstan is scaling its AI ecosystem from infrastructure to models, with Jensen Huang visiting Astana to meet President Tokayev and announce NVIDIA's investment in Kazakh-founded startup Higgsfield AI alongside new compute and energy partnerships.
NVIDIA 发布 Minerva，一个面向 AI 工厂的 AIOps 平台，通过整合网络、GPU、存储与工作负载的遥测数据，并利用专用 AI 模型构建动态知识图谱，实现异常检测、故障预测与根因分析，帮助运营团队从被动响应转向主动维护，最大化 GPU 产能与正常运行时间。
A repair shop reported receiving eight Nvidia GeForce RTX 5090 and RTX 4090 graphics cards with melted 16-pin power connectors within just two days, highlighting recurring durability issues with the high-power connector design.
Call of Duty: Modern Warfare 4 launches on NVIDIA GeForce NOW with RTX 5080-class ray tracing for Ultimate members, alongside in-app membership purchases on Fire TV Stick and additional new titles joining the library.
Nvidia announced the OpenShell Reference Architecture, a framework for building, containing, observing, and enforcing safe AI agents across enterprise infrastructure and edge systems throughout the full agent lifecycle.
At GTC Berlin, major European banks and financial firms showcased production-scale AI built on NVIDIA accelerated computing and open-weight models, including Revolut's PRAGMA foundation model, Lloyds' ATUM, and Starling Intelligence. Institutions like Santander, NatWest, and bunq are adopting NVIDIA's Nemotron and Kumo model families to build proprietary intelligence on their own data at lower cost.
NVIDIA expanded its Cosmos platform with new training tools, simulation checkpoints and Cosmos Data Factory blueprints to help developers build, evaluate and deploy world models for physical AI. Partners including Porsche, Volkswagen Group, ANYbotics, Microsoft Azure, Nebius and CoreWeave are already using the blueprints to accelerate autonomous vehicle and robotics development.
NVIDIA highlighted how European retailers are adopting its open models and physical intelligence technologies, with Shopify cutting inference costs by 90% and retailers like zooplus, Vestiaire Collective, and L'Oréal using digital twins and AI for product discovery, image processing, and research.
NVIDIA announced that organizations across Europe, including AXA, Vaisala Xweather, University of Manchester, and TotalEnergies, are deploying its Earth-2 open AI weather and climate models, with the new StormScope EU regional nowcasting model joining the family at GTC Berlin.
NVIDIA 和其欧洲合作伙伴宣布多项利用开放模型的公共服务项目，涵盖欧盟委员会的 AI@EC 平台、波兰的医疗影像与就业服务智能体系统，以及英国本地化数据建设。
Jensen Huang closed his GTC Berlin 2026 keynote with a Hermes Agent demo, using a single prompt to set up a customized Nous Research Hermes on NVIDIA RTX Spark, which then planned and simulated an obstacle course to train a Hugging Face Microduck robot policy before deploying it to the real world. Nous Research highlighted this as a showcase of open-source as a practical tool for chaining local and cloud models, devices, and tools.
NVIDIA 发布预训练关系型基础模型 Kumo Relational，可直接在企业数据库上通过单次前向传播完成预测，无需训练与特征工程；在 41 项任务与 15 个数据库上超越监督式关系深度学习及专家特征的梯度提升树管线。
gpu-unlock 1.0.0 发布候选版是一个 Linux 内核模块，通过在 NVIDIA 驱动绑定前解锁 RTX 3090 的 BF16 全速档位，免费实现约 15% 的 prompt processing 与 5% 的 token generation 提速，无需修改驱动或 UEFI。
A Linux kernel module hack reportedly unlocks 1.9x faster BF16 tensor math on NVIDIA RTX 3090 cards, boosting BF16 TFLOPS from 82 to 155 and cuBLAS GEMM performance, since NVIDIA ships the cards with fp32-accumulate tensor math at half rate.
At Nvidia's GTC in Berlin, Jensen Huang laid out a European vision centered on modular plug-and-play DSX Pod data centers, physical AI, and AI agents, while announcing robotics platforms and partnerships with VW and Mercedes.
记者 Natalie Fratto 在柏林 NVIDIA GTC 现场报道，分享了黄仁勋主题演讲的亮点之一——与 microducks 的互动环节。
NVIDIA 与 Hugging Face 推出 SimReady Central Library，提供超过 2,000 个模拟就绪（SimReady）OpenUSD 3D 资产，免费开放给开发者用于物理 AI 训练环境和 AI 工厂数字孪生。
NVIDIA在GTC柏林展示其本地AI硬件全线产品，包括RTX Spark、DGX Spark 64GB和DGX Station，并宣布多项新产品、软件更新及对Nemotron、DeepSeek、Qwen等新模型的首日支持。
Google Cloud and NVIDIA announced the general availability of A5X instances powered by NVIDIA Vera Rubin NVL72, scaling to up to 960,000 Rubin GPUs for agentic inference and large-scale reinforcement learning, with early customers including Ineffable Intelligence and WPP.