搜索“NVIDIA”

行官。 节目笔记 我们此前与 SF Compute 的对话 2020年石油期货为何跌为负值——EIA SF Compute 客户如何转售未使用的 GPU 容量 SF Compute 的 2.45 亿美元算力销售 Abel：SF Compute 的交割账本 CFTC：巴克莱与 LIBOR 操纵案 弹性 InfiniBand 分区 时间戳 0:00 — 为什么 NVIDIA 不能做这件事？ 0:25 — 欢迎回到 Latent Space 2:23 — 从卖一间房到经营一家酒店 3:47 — 物理交割：真正交付算力 6:00 — 流动性与 GPU 订单簿 8:42 — 交易非标准算力 10:00 — 中断和迁移工作负载 12:20 — 为什么硬件批次很重要 14:32 — 算力合同能延伸到多远？ 16:19 — 批处理与实时推理租约 17:58 — 算力价格的形态 20:33 — GPU 代际、规格与 SKU 23:30 — 物理交割与现金交割 27:00 — 新兴的算力市场生态 29:03 — 交付一个 GPU 集群需要什么 32:24 — 主权算力 33:26 — 为什么 NVIDIA

匹配字段正文

一定会遵守。如果你邀请德国朋友晚上7点来吃饭，我们会在6点58分到达。” “我们可靠、可预测、严谨,而且不会让人意外,”默茨说道,”简而言之,我们正是投资者最青睐的那种人。" 财政部长克林拜尔在向潜在投资者喊话时同样用了直白的幽默:” 我想要你们的钱。" 二人的推介看来都奏效了——峰会期间达成了数千亿欧元规模的投资承诺。 峰会期间，德国与美国科技巨头英伟达（Nvidia）达成协议，将深化双方在人工智能领域的合作，推动人工智能技术在德国工业中的应用。 在回答默茨提问“欧洲应如何监管AI”时，黄仁勋表示，如果AI产业无法蓬勃发展，“那就没什么可监管的。”图像来源: Fabian Sommer/dpa/picture alliance 值得一提的是，默茨在与黄仁勋举行的同台对谈中发问“欧洲应如何监管AI？”后者直言答道：“如果技术不存在，就无从监管。”黄仁勋认为，产业必须研发出安全技术，再建立具约束力的标准，最后才是政府与产业制定规范，如果AI产业无法蓬勃发展，“那就没什么可监管的”。 德国吸引外资优势与挑战并存 超过200名来自德国及国际大型企业、投资银行和投资基金的高管参加了此次峰会。周

匹配字段正文

迁移至HW3（特斯拉的第三代自动驾驶硬件）硬件，该方案让搭载HW3的车型也能获得HW4车型FSD迭代带来的提升，包含强化学习（RL）以及离线模型能力。 那么为什么特斯拉此次更新中强调向搭载英特尔芯片的Model S和Model X推送？因为Model S和Model X随着时间发展拥有搭载多个不同车机芯片的版本。 初代特斯拉Model S和Model X采用NVIDIA Tegra车机芯片。2018年起，特斯拉为Model S和Model X换用了英特尔Atom车机芯片，2021年起又换为AMD Ryzen车机芯片。 要注意的是，这里的芯片指的是车机芯片，这些车型的自动驾驶硬件还是HW3，不然是收不到FSD v14 Lite的升级的，特斯拉是用“英特尔芯片的Model S和Model X”代指这些老车型。 FSD v14 Lite在今年发布时仅面向使用HW3的特斯拉Model 3 和Model Y，后来确认会对Model S、Model X推送，而搭载AMD Ryzen车机芯片的HW3版Model S、Model X已经在今年9月下旬获得了推送。 在目前，已经有多位车主在社交媒体平台上

匹配字段正文

inese AI lab has faced scrutiny from governments and regulators across the globe since its R1 reasoning model shook the AI industry in January 2025, triggering a sharp sell-off in Nvidia shares. The U.S. accused DeepSeek and five other Chinese AI companies, including Alibaba and Moonshot AI, of using large-scale model distillation to extract information from American AI systems, according to a September report. In addition, OpenAI and Anthropic have separately raised concerns about DeepSeek’s al

匹配字段正文

s model serving across computing infrastructure. IBM, its Red Hat subsidiary and Google LLC are its lead maintainers. They co-founded the project last year with CoreWeave Inc. and Nvidia Corp. Other contributors include Advanced Micro Devices Inc., Cisco Systems Inc., Hugging Face Inc., Intel Corp. and Mistral AI SAS. The project offers organizations seeking more control over inference costs and infrastructure choices a way to reduce dependence on a single model provider, cloud or hardware platf

匹配字段正文

to have the technology removed. Over the summer, the US government considered steps against Chinese open-weight models , according to the New York Times. The options included sanctions, a trade blacklist, and rules on which models federal agencies are allowed to use. The push came after a wave of powerful Chinese models such as Kimi K3 from Moonshot AI. The government pulled back for now after Nvidia, Microsoft, Google, and Meta objected, while OpenAI and Anthropic had lobbied for restrictions.

匹配字段正文

nd this one thing might be controlled in three years, after Xi has spent years giving speeches saying we need to indigenize — I think the story Beijing told itself about importing Nvidia chips repeats here. “You don’t want to sell them to us at some point in the indefinite future? Fine. We’ll figure it out. We’ve got a crack team of engineers, we’re stealing ASML blueprints, see you on the other side.” As opposed to throwing down the gauntlet and forcing Trump to cave on something hard to cave o

匹配字段正文

路透社：英伟达已告知客户和合作伙伴，由于中国科技企业的需求远超计划产量，公司无法保证能够满足这些订单图像来源: Justin Sullivan/Getty Images/AFP （德国之声中文网）据路透社周二（10月20日）援引两名知情人士报道，中国科技巨头字节跳动、阿里巴巴和腾讯已向美国芯片制造商英伟达（Nvidia）订购了超过300万块新款RTX PRO 5500显卡，按标价计算总额超过350亿美元（约合310多亿欧元）。这款芯片性能不及英伟达最先进的产品，但符合美国对华出口管制规定。 报道称，英伟达已告知客户和合作伙伴，由于中国科技企业的需求远超计划产量，公司无法保证能够满足这些订单。上述消息人士因未获授权对媒体发言而要求匿名。 需求远超产能 或仅能交付一成 据路透社报道，一名消息人士透露，英伟达向合作伙伴表示，他们最终可能只能拿到所订数量的5%至10%。另一名消息人士称，英伟达表示2027年全球范围内的RTX PRO 5500产量最多只有300万块。两名消息人士说，来自中国客户的订单总量已超过600万块。 字节跳动、阿里巴巴和腾讯三家公司所订显卡的单价约为1.14万美元（约合1

匹配字段正文

JST-2 胜率）： · 角色生成：胜 Meshy 7.1（58.4%）、Tripo H3.1（73.2%） · 动画就绪度：胜 Meshy 73.8%、Tripo 88.9%，优势集中在手部绑定与变形质量 · 文本生成运动：胜 Cartwheel Swing 74.8%、HY-Motion-1.0（应该是 Hyper3D）80.0% · 视频生成运动：胜 Cartwheel 58.1%、DeepMotion 89.1%，手部追踪胜率高达 95% JST-3 的未来路线图：从“会动的 3D”到“学出来的物理” JST-2 中，角色的运动是学习到的动力学驱动的（这是它跳过绑定/蒙皮的原因），但物体之间的交互仍是规则模拟。JST-3 的目标是把学习动力学推广到物体交互，用学习到的物理替代基于规则的仿真。 最终愿景表述得很明确：为物理 AI 服务的“体验引擎”，智能体在模拟世界中通过行动和观察后果来学习。这本质上是对“机器人训练数据从哪来”这个问题的回答，与 NVIDIA Cosmos、Genesis、Google 的世界模型研究瞄准的是同一个终局，但入场角度是 3D 原生表示而非视频生成。

匹配字段正文

(1/6) 全新的 FastVideo FastH3-QAD-v2 和全新的 H3 NVFP4 VAE 解码器来了！🚀🚀 这整段 768p 视频是在两台 DGX Spark 上生成的，使用了 QAD-v2 和两款新解码器。8 步视频 + 音频，QAD 权重 💾 @NVIDIARTXSpark @NVIDIAAI @MiniMax AI https://t.co/90G6RcD2xx

匹配字段正文

Morgan Stanley rates Nvidia AI’s best risk/reward, though 91.7% of its 2027 revenue is exposed to infrastructure demand. MS research expects hyperscalers and AI cloud providers to spend $1.58T in 2027, up roughly 60%, with nearly $700B reaching Nvidia. That haul leans on 2 buyers, with SpaceX and Amazon Web Services modeled at about 54% of Nvidia’s 2027 hyperscaler revenue and 72% of growth. Morgan Stanley still calls the stock cheap at 15 times fiscal 2028 earnings, arguing analysts underestima

匹配字段正文

hNeuron that lets AI startups easily move their AI apps onto Amazon’s AI chip, Trainium. In doing so, Amazon is also taking a direct stab at the software advantage that has helped Nvidia become the default hardware choice for AI apps and spurred its rise as AI hardware provider of choice. TorchNeuron works with PyTorch, a free, open source toolkit that developers use to build, train, and run AI models. It lets AI apps written with PyTorch run on Trainium by changing just one line of configuratio

匹配字段正文

an agent appears to be making risky decisions. This idea of agents as untrustworthy insiders is taking hold across the broader industry, too. Microsoft was an inaugural member of Nvidia’s new Open Agent Safety Platform, which garnered the support of more than 100 other tech companies . The platform isolates agents in secure environments and can monitor and stop behavior through hardware, functioning something like a kill switch. This follows Microsoft’s announcement of the new Surface Laptop Ul

匹配字段正文

g-3.0-tts-0.5b Or build the 0.5b Docker image (torch 2.6, CUDA 12.4, transformers==4.46.3): docker build -f docker/Dockerfile -t ming-tts . The 0.5b environment has been tested on NVIDIA H800-80GB and H20-96G. For 0.5b, flash-attn is optional; inference falls back to PyTorch SDPA. Tiny uses a separate reference inference environment: Python 3.10.15, PyTorch and torchaudio 2.7.1+cu118 (CUDA 11.8 wheels), and transformers==4.46.3. Install the pinned dependencies from requirements-tiny.txt in a ded

匹配字段正文

中安装 SGLang： uv pip install --prerelease=allow sglang 接下来，启动您的模型： 快速入门: 运行您的第一个模型并发送请求。 示例手册: 选择您的模型和硬件，获取可直接运行的启动命令。 支持的硬件 SGLang 支持多种 GPU、TPU、NPU、CPU 和 Apple Silicon 平台。 平台 代表性硬件 NVIDIA B200/B300/GB200/GB300；H100/H200/H800/H20；A100；RTX 30/40/50 系列、RTX 6000 Ada / PRO 6000；DGX Spark、Jetson Orin AMD Instinct MI300X、MI325X、MI350X、MI355X Google TPU v6e、v7； SGLang-TorchTPU / SGL-JAX Intel（GPU / CPU) Arc / Arc Pro B 系列 GPU、Xeon CPU Apple Silicon 通过 Metal / MLX 的 Mac 华为昇腾 A2、A3、950PR/DT NPU 摩尔线程 MTT S5

匹配字段正文

nd the companies who’ve agreed to these contracts can’t cancel them except under some very limited circumstances. Its biggest customers are Microsoft, OpenAI, and what is probably Nvidia, though CoreWeave doesn’t disclose who its third-biggest customer is. Those three generated almost three-quarters of the company’s second quarter revenue. Microsoft and Nvidia are probably good for the money — at least, as long as it benefits them to be good for the money. (I suspect they realize that if CoreWea

匹配字段正文

y chips will continue to surge through 2028. And for years the company watched its domestic rival SK Hynix pull ahead in the global market for the high-bandwidth models that power Nvidia’s AI accelerators. (Meanwhile, the Taiwanese chipmaker TSMC has dominated the foundry business, manufacturing chips that other companies design.) The Pyeongtaek build-out is Samsung’s attempt to reclaim its lead on memory chips. Although the heat can be brutal and the work zones are crowded, Hye-rin has no plans

匹配字段正文

party tools, and partner testing. - Claude is preparing a Personal Finance feature, formerly codenamed "Claude Money", that analyzes bank data through Plaid and builds dashboards. NVIDIA 🔥: - NVIDIA GTC Berlin runs Oct 20–22, with Jensen Huang's keynote at the Tempodrom on Oct 21. MINIMAX 🔥: - MiniMax teased the upcoming MiniMax-M3.1-Flash. It is now being rolled out on the APIs. OPENCODE 🔥: - OpenCode said a distillation attack drained the OpenAI, Anthropic, and Google API balances of about 1.4

匹配字段正文

Nvidia on Thursday launched new open-weight models designed for always-on agents and edge computing. The Nemotron 3.5 Super VL is a large, 120-billion-parameter model geared for agents used by the enterprise. While it uses fewer parameters than the Nemotron 3 Ultra model, Nvidia says it can deliver similar accuracy in instruction-following and tool-use tasks. At the other end of the spectrum, the company also launched two lightweight Bolt models. These are intended to run locally on the new AI P

匹配字段正文

speaks through it, with face on its visor. It stands 5'5", weighs 176 lb., and rolls on a three-wheel omnidirectional base. It handles objects with two-finger grippers and runs on Nvidia’s Jetson Thor. Why it matters: Plim says it has been working in real homes since 2024, starting in elderly care, and is now pitching to both families and care organizations. On its website, bedtime stories from faraway loved ones share the spotlight with household chores, but it still comes without a price, deli

匹配字段正文