跳到正文
10月5日周一
10月3日周六
  1. NVIDIA26

    感谢 The Korea Society 向 @JensenHuang 颁发 2026 年 Van Fleet 奖,并庆祝美韩伙伴关系。 我们与韩国合作已超过 25 年,我们为共同取得的成果感到自豪。期待未来更多的合作。

    AI 生成摘要 · 以原文为准

    引用The Korea Society@koreasociety

    The Korea Society thanks all distinguished guests who joined us for an unforgettable evening on Sept 28 celebrating the U.S.-Korea partnership and 2026 Van Fleet Awardee, @JensenHuang (@NVIDIA). Our sincere thanks to our generous sponsors who made the 2026 Annual Gala possible.

  2. Fabricated Knowledge33

    笑死,这渲染搞得我火大——兄弟,我们根本不是在 PCIE 上用显卡的

    AI 生成摘要 · 以原文为准

    引用Bearly AI@bearlyai

    Someone used Claude Opus 5.5 to make a 3D animation of Nvidia Blackwell GPUs (from server down to an atom). Took 1 hour and full animation “runs from single HTML file, directly in browser. No vid editor. No pre-rendered 3D sequence. Just browser-based experience.” Very cool.

10月2日周五
  1. NVIDIA Newsroom64

    OpenAI 发布 GPT-6 Astra Ultrafast,NVIDIA Blackwell GPU 加速下 token 生成速度最高提升 8 倍

    GPT-6 Astra Ultrafast 现已在 OpenAI API 及面向符合条件的 ChatGPT Work 与 Codex 用户开放,运行在 NVIDIA Blackwell GPU 上。

    AI 生成摘要 · 以原文为准

    关注理由:OpenAI 基于 NVIDIA Blackwell 推出 Ultrafast 推理模式,展示了模型厂商通过自研推理优化榨取 GPU 硬件性能的路径,对算力平台软件生态有意义。

10月1日周四
  1. NVIDIA Newsroom43

    NVIDIA 自豪地加入美国技能行业联盟(Alliance for America’s Skilled Trades)。 AI 正在推动一代人一遇的基础设施建设浪潮,而美国需要熟练劳动力来建设它。 该联盟的新报告发现,到 2035 年,每年将需要填补 170 万个技能行业岗位空缺。

    AI 生成摘要 · 以原文为准

    引用Jim Farley@jimfarley98

    At @Ford, we care a lot about the Essential Economy, which is the workers who build, fix, and maintain the physical world around us. These skilled trades are the backbone of American manufacturing and we need solutions for closing the skills gap, and providing the education and resources that will power the productivity and growth of the #EssentialEconomy. Across industries and public and private sectors, we all need to work together. That's why this summer, Larry Fink from @BlackRock, Ruth Porat from Alphabet Inc. & @Google, Linda Hubbard from @Carhartt, and I launched the Alliance for America's Skilled Trades. Among other goals, we wanted to build a shared, trusted foundation of data on the state of skilled trades across America. So today, in partnership with Jobs for the Future (JFF) and The Burning Glass Institute, we’re officially publishing — The State of America's Skilled Trades: A National Report. Take a look: https://skilledtradesreport.com The report examines the forces shaping supply and demand across occupations and regions, to show where workforce systems are working well, where they're under strain, and where focused action can make the biggest difference. Alongside the report, we're launching America's Skilled Trades Dashboard — a first-of-its-kind interactive tool that makes the data accessible in one place. It breaks down the data by occupation and state, making it easy to explore local labor market conditions, identify where the need is greatest, and compare progress across regions. The dashboard is available at https://skilledtradesreport.com We also set out to grow this movement. Today, we're welcoming 14 new partners - @ATT, Barton Malow, @blackstone, @CintasCorp, @DEWALTtough, Dycom, @Hersheys, @ManpowerGroup, @Meta, @Microsoft, @MilwaukeeTool, @nvidia, @Verizon and @Waymo. Each has already stepped up individually. Now we're stepping up together. You can read more about today’s news here: https://www.fromtheroad.ford.com/us/en/articles/2026/alliance-for-americas-skilled-trades-expands-national-report #FordAccelerate #Accelerate2026 #SkilledTrades

  2. NVIDIA Blog31

    NVIDIA 开放 2027–2028 学年研究生奖学金申请,每人最高 6 万美元

    NVIDIA 开放第 26 届研究生奖学金计划申请,面向从事 AI、机器学习、自动驾驶、机器人等研究的博士生,每人奖金最高 6 万美元。申请者须已完成至少一年博士阶段学习,并在 2027 年暑期到 NVIDIA 研究办公室线下实习,截止日期为 2026 年 10 月 30 日。该计划自 2002 年启动以来已颁发超 220 份资助,总额约 800 万美元。

    AI 生成摘要 · 以原文为准

9月29日周二
  1. SemiAnalysis72

    SemiAnalysis解析GLM-5.3稀疏注意力如何影响HBM内存占用

    SemiAnalysis撰文分析GLM-5.3的稀疏注意力与DSA架构对HBM内存和推理成本的影响。文章指出top-k选择仍需完整上下文驻留HBM,稀疏注意力并不消除内存容量瓶颈,SGLang的HiSparse通过将KV cache按LRU策略在HBM与主存DRAM间分层换入换出并跨层重叠加载,在高并发长上下文下维持95%以上缓存命中率。

    AI 生成摘要 · 以原文为准

9月28日周一
  1. NVIDIA55

    NVIDIA 联合超过 100 家行业伙伴推出 NVIDIA Open Agent Safety Platform,整合 OpenShell 与 Sentry,定位为安全智能体系统的开放信任层。黄仁勋表示,AI 的潜力只有在人们确信其被安全、负责任地构建和部署时才能实现,安全是赢得信任的方式。

    AI 生成摘要 · 以原文为准

    引用Jensen Huang@JensenHuang

    Today, with over 100 industry partners, we introduced the NVIDIA Open Agent Safety Platform, bringing together OpenShell and Sentry. Artificial intelligence is extraordinary technology that will advance discovery, productivity, security, health, and prosperity for generations to come. But its full promise can only be realized when people have confidence that AI is being built to be safe and deployed with wisdom and responsibility. This is bigger than a single product. It's the beginning of an open ecosystem to build the trust layer for safe agent systems. Together, we are building the foundation of the AI economy. Trust and innovation are not in conflict. Safety is how trust is earned. We must build not only the most capable AI, but the most trusted AI, so that this extraordinary technology can realize its enormous promise for the world. https://nvda.ws/4hcoq7m

9月27日周日
  1. Dylan Patel53

    SemiAnalysis 的 Dylan Patel 发推称 Rubin HBM 降规格(Despec)等于 Nvidia 减料,并引用一篇讽刺文:GB200 NVL72 约 400 万美元、GB300 约 500 万美元(重约 1580 kg),Vera Rubin NVL72 已被报至最高 880 万美元,机柜重量受数据中心地板承重限制而价格不受限,$/kg 每代约 1.75 倍增长,外推 2030 年与可卡因批发价出现交叉。内容为戏谑观点而非测算结论。

    AI 生成摘要 · 以原文为准

    引用Dylan Patel@dylan522p

    Fentanyl Grade Compute: Why a GB300 Rack Will Out-Price Blow by Weight in 2030 A GB300 NVL72 weighs roughly 1,580 kg fully populated and recent purchase orders put it at $5M per rack. That is $3,165/kg. Strip out the 1.5 tons of busbar, manifold and coolant and the GPU packages alone are well into gold territory, but we're pricing the rack, because that's what you actually take delivery of. Where that sits on the illicit commodity curve today ($/kg): $2,400 Cannabis flower $3,165 GB300 NVL72 $3,500 Fentanyl $28,000 Cocaine $65,000 Heroin $138,000 Gold So today NVIDIA ships a product that is denser in value than weed, roughly fentanyl-grade, and still an order of magnitude short of cocaine. For now... GB200 NVL72 was $4M, GB300 is ~$5M, and Vera Rubin NVL72 is already quoted at up to $8.8M at essentially the same rack mass. Rack weight is constrained by the datacenter floor loading; rack price is constrained by nothing. That's ~1.75x $/kg per generation with HBM, CoWoS and power all supply-limited through the decade. Cocaine: Colombia's new president has pledged a hard-line security crackdown with $1B in US aid behind it Every prior crackdown has consolidated the industry into fewer, better-capitalized operators with better logistics and higher yields per hectare. Consolidation is deflationary. We model US wholesale drifting from ~$28k/kg toward ~$12k/kg by 2030 as the supply chain professionalizes. **Crossover: 2030.** Our extrapolated NVL72-class rack hits ~$17k/kg while wholesale cocaine falls to ~$12k/kg. At that point the rational cartel pivots to smuggling racks, except a rack has a fixed 500+ kW power draw, a 3,300 lb forklift requirement and an export-control regime that is actually enforced. Cocaine has none of these problems, which is why it will remain the superior product for anyone without a substation. Key risks to the thesis: US retail coke is still ~$60–200/g, i.e. $60k–200k/kg so the crossover only holds at wholesale. Retail compute (H100 hour on a neocloud) is also marked up, so we consider this an apples-to-apples wholesale comparison. Full model available to Coke Research subscribers.

9月25日周五
9月24日周四
  1. SemiAnalysis69

    SemiAnalysis 发布 ClusterMAX 3.0 GPU 云评测:77 家厂商分级,Nebius 升入铂金级

    SemiAnalysis 发布 ClusterMAX 3.0 GPU 云评测,覆盖 77 家被测厂商、323 家市场观察对象(上一版为 209 家),并采访了 200 多位终端用户。Nebius 加入 CoreWeave 位列铂金级,Google Cloud 与 Oracle 列黄金级,Azure 降至白银级,Fluidstack 转为 Unavailable,Crusoe 降至青铜级。

    AI 生成摘要 · 以原文为准

9月23日周三
  1. IonQ69

    IonQ 宣布其 IonQ Superion 256 系统将成为 NVIDIA 加速量子研究中心(NVAQC)首个本地部署的量子计算系统。该系统将通过 NVIDIA NVQLink 直接连接一台 NVIDIA GB200 NVL72 系统,工作负载由开源的 NVIDIA CUDA-Q 平台编排,目标是把量子处理器扩展为加速量子超级计算机。

    AI 生成摘要 · 以原文为准

    关注理由:IonQ 量子系统将首次部署在 NVIDIA 量子研究中心内并与 GPU 集群直连,为混合量子-经典计算的硬件协同提供具体落地方向。

9月22日周二
9月21日周一
  1. NVIDIA54

    Einride 宣布与 NVIDIA DRIVE 达成战略合作,在高速公路和郊区道路上加速自动驾驶卡车运输。新一代 Einride Driver 将基于 NVIDIA Hyperion 平台构建,以更具资本效率的方式扩展 Einride 货运网络的自动驾驶运力。NVIDIA 官方账号转发了这一合作消息。

    AI 生成摘要 · 以原文为准

    引用Einride@einrideofficial

    Einride enters a strategic collaboration with @NVIDIADRIVE to accelerate autonomous trucking on highways and suburban roads. Building the next generation of Einride Driver on NVIDIA Hyperion platform, supports capital efficient scale of autonomous capacity across Einride’s freight network. $ENRD

9月20日周日
  1. NVIDIA Newsroom37

    NVIDIA CEO @JensenHuang 做客 CBS Sunday Morning,与 @jolingkent 畅谈 AI 发展、安全,以及行业发展应该有多快。 观看对话:

    AI 生成摘要 · 以原文为准

    引用CBS Sunday Morning 🌞@CBSSunday

    The Atlantic recently called Nvidia founder and CEO Jensen Huang "the most powerful man in AI." He talks with CBS News' @jolingkent at Nvidia headquarters in Santa Clara, Calif., about the latest developments in artificial intelligence and the ongoing debate over safety and regulations. https://cbsn.ws/476ihUf

9月19日周六
9月18日周五