NVIDIA祝贺StepFun发布Step 5 Preview开放权重模型
StepFun宣布Step 5 Preview开放权重,采用总参数量600B、每个token激活27B的稀疏MoE架构,支持1M-token上下文窗口和原生文本、视觉输入。
推荐理由:StepFun公布Step 5 Preview开放权重,并列出模型规模、上下文窗口与输入形式,NVIDIA则补充了GB300上的运行速度和NVFP4口径。
X(推特)上的财经、科技与宏观账号
StepFun宣布Step 5 Preview开放权重,采用总参数量600B、每个token激活27B的稀疏MoE架构,支持1M-token上下文窗口和原生文本、视觉输入。
推荐理由:StepFun公布Step 5 Preview开放权重,并列出模型规模、上下文窗口与输入形式,NVIDIA则补充了GB300上的运行速度和NVFP4口径。
We want to support the mathematics community in turning progress in AI into deeper human understanding and new discoveries. To support that goal, we are contributing $5 million to a new Simons Foundation initiative to help mathematicians examine, understand, and build on results 引用 @SimonsFdn:How c
After an ongoing deployment in Barcelona, @CocaColaEP is scaling Spot globally. Spot is already up and running in East Kilbride, Scotland, locating air leaks and automating condition monitoring rounds. https://t.co/vKmQnIcpRc
Nemotron 3.5 Super VL 将文本和视觉理解整合到一个模型中,减少在独立文本模型与视觉模型之间传递上下文的需要。აგენტ可能需要阅读文档、解读图表并搜索网页来完成任务。
英伟达推出开放模型 Nemotron 3.5 Super VL,将推理和工具使用与文本、图像、文档及视频理解结合,用于长时间运行的 AI agents。该模型总参数量为 120B,激活参数量为 12B。
Muse Spark Open 今日可通过 Baseten、Fireworks AI、Groq、Modal、Nebius、OpenRouter 和 Together AI 的 API 平台使用。
Muse Spark Open 针对两台搭载 Apple M5 Ultra 的 Mac Studio 进行了优化,并支持最新 MLX 框架。它也可运行于 SGLang、vLLM、TensorRT-LLM、Hugging Face Transformers、llama.cpp 和 Unsloth 等推理引擎。
Sferical One 这台由 NVIDIA 提供算力支持的共享 AI 超级计算机举行启动庆祝活动。黄仁勋与 Wallenberg Investments 的 Marcus Wallenberg、Spotify 的 @eldsjal 及 Sferical AI 的 Jenny Nordlöw 共同出席。
🔥 引用 @poteto:hey @bot please preorder the new Tesla Roadster for me immediately, thanks
马斯克称,SpaceX与英伟达共同设计的Star Mind计算机性能远好于多数人的预期,并计划明年进入量产。马斯克还表示,与英伟达工程团队合作令人愉快。
When Konstantin called me way back then, I said I had no idea what it would look like or when it would be made, but he still had the faith to put down a deposit ❤️ 引用 @konikaze:@elonmusk My original Roadster still runs 20 years later. Battery still fine! Reserved the new one first in 2013. Thirteen
New voices. Better accents. And less "checking..." GPT-Live-1.1 brings these updates to Voice in ChatGPT, Codex, and dots, with new voices in French, German, Hindi, Mexican Spanish, Korean, and Japanese. Rolling out today to all users. Coming soon to the API. https://t.co/L6K8ZUC1xm
When the brand synonymous with car safety for almost 100 years teams up with AI, amazing things happen. 🚗 We’re expanding our partnership with Volvo Cars to advance safe automated driving: Volvo’s safety expertise will come together with the NVIDIA Hyperion, Halos OS and https://t.co/8pszLgH2tg
Meta发布开放权重模型Muse Spark Open,称其为迄今能力最强的美国开放权重模型,面向编程和智能体任务。模型总参数量为816B、激活参数量为42B,支持1M token上下文窗口;NVIDIA称该模型在NVIDIA GB300 NVL72上训练。Meta还制作了针对Blackwell Ultra优化的NVFP4检查点,原文随后提到“over 50K”,但未提供完整表述。
英伟达推出 Nemotron 3.5 Bolt,面向 RTX Spark 上运行的本地 AI 智能体。该模型参数量为 74B,可在 24GB 内存中运行。
英伟达宣布扩展 Nemotron 开放模型家族,加入 Nemotron 3.5 Super VL。该模型结合视觉理解、推理与工具使用,面向 AI agents;引用内容称其支持文本、图像、文档和视频理解,总参数量为 120B、激活参数量为 12B。开放模型可供用户定制、部署和构建。
NVIDIA 推出 Nemotron 3.5 Bolt 24GB,称其可在 24GB 内存中运行。该模型包含 74B-A9B 参数,面向 RTX Spark 优化,可与其他应用同时运行。
AI has a lot to learn from the Nordics. Tonight, Jensen celebrated the builders combining their expertise with NVIDIA technology to tackle challenges in manufacturing, medicine, and beyond. Thanks for an inspiring evening, Stockholm. 📍Next stop: Berlin. https://t.co/ydtjTOILgS
Starlink is really that good 引用 @SawyerMerritt:New opinion article about @Starlink from a 7-year Delta Airlines Diamond Medallion status member (their highest loyalty tier): "After flying a United flight in July that included Starlink, it was enough reason for me to switch. It was amazing, like be
Makes sense 引用 @tylerpalmer:imagine an Amazon competitor that is: 1) agent first - built for personal agents 2) america first - sells only products from usa might actually be a huge market for this.
Learn about the Rosalind Academic Plan, check your lab's eligibility, and find out how to apply: https://t.co/xw86gMoj5C
GPT-Rosalind is built to support that work: helping researchers reason across scientific literature and biological data, with scientists bringing the expertise and judgment to evaluate the evidence. Eligible labs can receive up to 12 months of sponsored access to OpenAI tools,
We're expanding access to GPT-Rosalind for academic research. The new Rosalind Academic Plan helps scientific labs bring AI into the work of understanding biology, exploring ideas, and deciding what to test next—with sponsored access for eligible labs at U.S. universities. https://t.co/1O702wdc5f
Working out how best to use AI in mathematics research—and in science more generally—will take time, effort, and resources. We’re eager to support mathematicians as they drive scientific progress and define the future of their field. We’ve committed $5 million of funding, 引用 @SimonsFdn:How can the
𝕏 is the global town square. No place better to reach the world fast! 引用 @cb_doge:Rockstar Games posted its GTA 6 promo on YouTube and 𝕏. Just 90 minutes later: 𝕏: 7.5 million views YouTube: 1.6 million views Nearly 4.7 times as many reported views on 𝕏. https://t.co/ZGZhnIdg05
NemoClaw was just named one of TIME’s Best Inventions of 2026! 🦞 So proud of the team and shoutout to everyone building safer AI agents with us. Blurb on @TIME: https://t.co/EY5S9htCn5 https://t.co/G3vo2N3RPP 引用 @TIME:These are creations changing how we live, work, play, and think about what’s p
We’re proud to share that Meta Omnilingual Automatic Speech Recognition (ASR) has been named one of @TIME’s Best Inventions of 2026! Meta Omnilingual ASR is a suite of models providing ASR capabilities for over 1,600 languages, including 500 low-coverage languages never before https://t.co/pQjBPvE5
Grok @Bot improves significantly every day 引用 @jumperz:i was about to pay for a small Codex subscription just for computer use.. honestly, I didn't expect Grok bot's computer use to be this fast... I've been more and more impressed with it every day lately. for day to day stuff, nothing else I've
Never seen anything grow as fast as @Grok @Bot! And there is no free tier, which means people are deriving real value in excess of what they pay. As the saying goes, you get what you pay for. 引用 @spectnfa:HOLY SH*T, JENSEN HUANG SAYS GROK BOT USAGE IS SKYROCKETING, AND ELON IS DOING SOMETHING AL
We’ve committed to having evaluators assess our safety practices. We're in discussions with multiple organizations, and have now given employee-like access to a team from @METR_Evals. In this post, we describe how we work with third-party evaluators: https://t.co/8CLQnMVg4X
NVIDIA表示,FLUX 3 Image Klein 5.6B在NVIDIA GB200 NVL72上训练,并可在RTX GPU上运行。NVIDIA还提到该模型有NVFP4变体;引用的@bfl_ai称其为开放权重模型,拥有56亿参数,可在单张消费级GPU上运行,并支持原生4K、边界框和像素级编辑。
Shanu Mathew称,燃料电池(BE)的标示成本约为5700美元/千瓦,较每千瓦2000至3000美元的燃气基准高90%至185%。他表示,35%的税收抵免及额外优惠可降低成本,但基础成本仍高;引用内容称,Google与AEP签订20年电力供应协议,涉及AEP一座耗资40亿美元、装机702MW的燃料电池电厂。
Nvidia副总裁Rod Evans在Vertiv的EMEA活动上表示,AI数据中心消耗的电力中,只有60%–70%用于有效AI工作。TechRadar称,他还拆解了一座用电负荷为1GW的数据中心的用电分配,但正文未提供后续明细。
财经博主转述并质疑Gartner关于新云服务商的预测,询问其量化依据是否可信。引用内容称,Gartner预计到2030年80%的新云服务商将停止独立运营;2026年AI服务商预计将在AI基础设施上支出近1万亿美元,并预测2027至2030年再投入6万亿美元。
Chicken McNuggets were a hit within weeks, but the story behind them spans decades. It began with a 1948 contest to breed a bigger chicken and ended with a then largely unknown @RayDalio, who found a way to hedge a chicken by hedging its lunch. Order a box today and you'll get https://t.co/GYN3dkc
Interesting tension across the AI trade this week. TSMC and ASML delivered strong numbers, yet the Anthropic run-rate debate and the CoreWeave selloff have investors questioning the economics underpinning the broader infrastructure buildout. >TSMC reported $46.7B of quarterly
AI infrastructure 👀 "*Every* open source inference cloud I am aware of is growing faster both sequentially and YoY than either Anthropic or OpenAI. And their combined compute spend is becoming material." 引用 @GavinSBaker:TLDR; we will probably never get another ARR number from Anthropic, which is
$ASML 3Q'26: 2030 Arrived Four Years Early full piece: https://t.co/bQ9T1PmUb1 https://t.co/hsnwP31qMx
NEW: Episode 3 of the Power On podcast is out! We break down Apple’s smart home launch, go hands-on with the HomeView, and debate the new Apple TV & HomePod mini. Plus, exclusive details on Apple’s upcoming touchscreen Mac launch, OLED iPad mini and more. https://t.co/yClml58ymf
YouTube: https://t.co/zcEOSWQgFu 引用 @markgurman:NEW: Episode 3 of the Power On podcast is out! We break down Apple’s smart home launch, go hands-on with the HomeView, and debate the new Apple TV & HomePod mini. Plus, exclusive details on Apple’s upcoming touchscreen Mac launch, OLED iPad mini and m