NVIDIA
Designer of the GPUs that power most modern AI training and inference. Central to the AI compute supply chain, frequently cited in crypto news around AI-narrative tokens, decentralized GPU networks, and tokenized compute.
新闻
. @RaoulGMI reveals the "universal code" he believes explains why Nvidia, AI, and crypto all keep winning "The entire universe, from atomic structure to solar system structure, converts units of energy into units of intelligence. We're part of that process. AI is part of that process. Blockchains a
JUST IN: Nvidia back above $5.5 trillion market cap
One of many great conversations at #GTC Taipei 2026! SK Group Chairman Chey Tae-won and @NVIDIA CEO Jensen Huang connected at the @SKhynix booth, highlighting the collaboration driving the next chapter of AI. Take a look at the moments from the show floor 📸 #SKhynix #NVIDIA
JUST IN: Nvidia CEO says he expects to return 50% of free cash flows to shareholders
Saylor Sells BTC - Market Panics 0:00 Saylor Sells Bitcoin 2:45 Bitcoin Charts 7:30 ETH SOL Updates 10:15 Stocks Generational Run 11:15 Hype 12:15 Software Stocks 13:55 Copper 14:45 AMD 15:15 New Nvidia Shill @saylor @piovincenzo_ @decodejar @shahh @andyyy @randgroup
Premarket movers: Mag 7 are mixed with Alphabet down 2.7% after raising $80 billion through a package of equity offerings, including a deeply discounted private placement with Berkshire Hathaway and a $40bn ATM ovvering (Nvidia +1.5%, Meta +0.5%, Tesla flat, Apple -0.1%, Amazon -1.6%, Microsoft -2.
英伟达今天宣布推出光电一体化的网络交换机：NVIDIA Spectrum-X™ 和 NVIDIA Quantum-X。 从光模块路线看，NVIDIA 的硅光 / CPO 交换机代表的是一条很清晰的升级路径： 传统可插拔光模块 → LPO 线性可插拔光模块→ 硅光模块 → CPO 光电共封装。 传统可插拔光模块是现在主流。交换芯片在交换机里面，光模块插在前面板，电信号要从 ASIC 经过 PCB、SerDes、DSP，再进入光模块变成光信号。优点是成熟、标准化、方便更换；缺点是 800G、1.6T 之后功耗、延迟、信号损耗和散热压力越来越大。 LPO 是过渡方案。它仍然是可插拔模块，但尽
We have worked with @nvidia to integrate their official Agent Skills catalog into the Hermes Skills Hub. These skills teach your agent how to use CUDA-X libraries, Omniverse and Physical AI workflows, NeMo training and inference tools, and other platform components.
NVIDIA JUST PUT A GAMING PC, AN AI WORKSTATION, AND A GPU SERVER INTO A LAPTOP IF THIS THING IS REAL, THE INDUSTRY IS ABOUT TO GET WEIRD
My GOAT Kimi-K2.6 is running at home finally Nvidia's NVFP4 + 47% Prune - 41.1% on terminal bench 2.0 - 5 of 89 samples caused repetition degeneracy - 39 tok/s decode - 950 tok/s prefill - 4x 6000 I still need to do some model surgery but we finally have something usable.