NVIDIA

4NE1 Gen 4 is already making its mark on the world stage. 🤖 Today, we unveiled the next generation of our humanoid robot on two major stages. First, at the opening of the TUM RoboGym powered by NEURA in Munich, built in collaboration with…

At GTC Berlin, Nvidia launched OpenShell: Vera Rubin NVL72 + BlueField-4 + Sentry to isolate and kill misbehaving AI agents in milliseconds. Dell, Cisco, HPE, Lenovo and Supermicro are building to it. Agent runtime is becoming a rack SKU.…

Nvidia case study: Shopify’s post-trained Nemotron 3.5 Lightning GraphQL subagent beat a frontier model on an internal bench while cutting serving cost ~90% and answering ~7× faster. Open weights plus post-train is the cost curve.…

A new step for Agility Digit 5. 🤖 @agilityrobotics is adopting NVIDIA Thor to simplify onboard compute and building on its NVIDIA Halos for Robotics collaboration, bringing compute and safety development together as it prepares for…

Jensen Huang just made some VERY interesting comments about the ongoing financing concerns surrounding neoclouds like $NBIS and $CRWV. One of the biggest concerns surrounding these companies has been their ability to secure financing at…

‘Jensen’s nightmare’: Huawei, DeepSeek narrow AI chip ecosystem gap with Nvidia DeepSeek is likely to rely ‘primarily’ on Huawei’s ‘already mature and well developed’ technology stack as domestic semiconductor industry gains steam at…

Nvidia CEO to deliver keynote speech as his firm faces investor scrutiny over AI The conference, focusing on sovereign AI, automation and infrastructure, comes days after Jensen Huang’s meeting with German Chancellor…

Someone posted the proof yesterday and it's been sitting with me since. Higgsfield's new Genjutsu Lite model, going by screenshots of their own API docs, is Alibaba's Wan 3.0 wearing a different name. Genjutsu is Seedance 2.5. Same…

GTC Dev Community Berlin Recap: From Open Models to Physical AI https://t.co/iNcjNOMiYC

Digit 5 will standardize on NVIDIA IGX Thor for its onboard compute, expanding our collaboration with @NVIDIA as we prepare Digit 5 for broader commercial deployment. Read the details: https://t.co/tXZ5TO8rFu https://t.co/5MtJdcqsha

mimic demonstrates real-time teleoperation of its dexterous robotic hand using MANUS gloves, live from NVIDIA GTC in Berlin. Video Credit: @ManusMeta #engineering #technology https://t.co/jbzB1Ck5XO

MicroDuck looks like a cute desk toy. The hardware tells a different story. At just 25 cm tall and around 780 g, this small robotic duck packs a surprising amount of engineering into its body. Here’s what’s inside: • 15 motors controlling…

Culture is the moat. Who has the most diverse factory in the world? Talked to a robotics entrepreneur today who explained this process to me while standing out front of the Physical AI event at Fort Mason. Stealth Mode. Building a robot…

A coding model that runs at 125 tokens a second on a MacBook and beats most open models under 50 billion parameters on SWE-bench was built largely by another AI system that picked its own architecture. That's Flint, out this week from…

Your #Yocto image boots on #Jetson, but what if NVMe or USB still isn't working? Using #reComputer Super J401 (@NVIDIARobotics Jetson Orin NX/Nano), we walk through matching the hardware config, building and flashing the image, and…

🤝 Physical AI for industrial production: Schaeffler and Wandelbots have entered into a technology partnership. Together, we will further develop the open Wandelbots NOVA platform for industrial and humanoid robotics. ➡️ Press release:…

エヌビディア、倉庫ロボットの学習・検証を統合-https://t.co/XcPED2iekR エヌビディアはロボット開発基盤「NVIDIA Isaac」を拡張。ヒューマノイドの学習・検証環境や安全技術を強化し、Agile RobotsやDexory、KIONなどが倉庫ロボットの開発や安全性向上に活用する。 https://t.co/Kxj3BcCWrG

According to Korean media, Samsung Electro Mechanics just signed the biggest substrate contract in its history, and the customer is $NVDA. The multibillion dollar deal covers the substrates that sit under Nvidia's AI chips and networking…

Meet the IP66 #reComputer Rugged Series — built for demanding industrial automation and outdoor deployments. Watch the full video on YT: https://t.co/ERrn8RWs3J Explore specs here: https://t.co/nIOASMkFML 🔹 @NVIDIARobotics Jetson Orin…

How do you teach a humanoid robot to walk without programming every move? Reinforcement learning step by step 1. Build a simulation (MuJoCo or NVIDIA Isaac Lab) with joints motors sensors and a physics model. 2. Define the task: walk…

NVIDIA、欧州で自動運転基盤を貨物車へ展開-https://t.co/ia2nCBCdoq エヌビディアは、欧州の自動運転開発で車載基盤「NVIDIA Hyperion」の採用拡大を発表。Einrideが自動運転貨物トラックに活用するほか、安全性評価の新枠組みも導入し、商用化を支援する。 https://t.co/TTD6Nh6gFZ

Goldman Sachs: NVL8 28x more compute compared to A100 for 3.7x more power. The result? 87% less power per petaflop $NVDA

Samsung's 12-high HBM4E cleared Nvidia qualification at end-September, per Hankyung. 48GB per stack, 16Gb/s per pin — over 20% faster than HBM4, 16% better efficiency, on a 4nm logic base die. No contracts or volumes disclosed. The base…

Wells Fargo: We currently model Nvidia shipping ~23gW of total AI infrastructure capacity in CY27/FY28, or an average of 5.6GW per quarter. $NVDA

This.... https://t.co/AAwQpUuiRN

Nvidia Moves to Back GPU Cloud Provider Firmus After Investors Balk at Its IPO https://t.co/MfTGiSYHY5

Three teams in one month have pushed LLM decode forward with megakernels: Inferact on Google TPUs, AMD with TileRT and FlyDSL, and now NVIDIA on GB300, now live on InferenceX. We're excited to see these gains and dig into the engineering…

Europe 💚🤖 https://t.co/5oM80ZUCl3

BofA: TrendForce states Nvidia's AI-server share moderates from 66.9% in CY26 to 62.6% in CY27 as “Others” rises from 26.0% to 29.6%. Custom ASICs increase from 26.1% to 29.6% of AI server shipments, led by hyperscaler programs. $NVDA…

Most people just saw a humanoid robot throwing kicks and flipping off cars. But the H2-X has some technical details worth looking at. Here’s what the footage and the available specifications tell us about Unitree’s Kung Fu Legend Guardian.…

With NVIDIA GTC underway, it's been a busy day in robotics; here's what's interesting across the industry today https://t.co/Vcr8qGUQOx

Why is nobody talking about this? ToddlerBot is a 0.56 m, 3.4 kg humanoid robot built at Stanford by Haochen Shi, Weizhuo Wang, Shuran Song and C. Karen Liu. Its goal is to make robot learning cheap and easy to reproduce. The full hardware…

$SPY rose about 0.5% today to around $773, bouncing from an early low of $768.41 as chip and AI stocks led the market higher. Here are the top stories you need to know. 1. $SMCI jumped about 10% after raising its fiscal first quarter…

‘The Fix Is In’ in AI. Musk, Nvidia, OpenAI & AI Agents. ARD #189 …a Tesla-SpaceX merger, Nvidia on water and power, OpenAI on Google’s cloud, and the AI agent race. $NVDA $TSLA $SPCX Full writeup and source links: https://t.co/LecVwJHf8A…

Generative Bionics and AMD are building a Physical AI data factory Generative Bionics is partnering with AMD to build a 5,000㎡ Physical AI Data Factory in Italy. The first facility is scheduled to begin operations in Q4 2027. The goal is…

Walked out of the NVIDIA keynote at GTC Berlin still chewing on the simulation piece! 🇩🇪 Physical AI is going strong. The world will be full of robots very soon. One of the most important things to create better robotics applications is……

Another major shortage is starting to develop across the AI supply chain. We're talking about multilayer ceramic capacitors, or MLCCs, which are tiny components used in everything from AI servers and GPUs to smartphones, networking…

By pairing unified driver architecture with NVIDIA JetPack 7.2, @AVerMedia and Stereolabs streamline edge vision deployment, delivering optimized memory usage, hardware compatibility, and faster production timelines for autonomous robotics…

Wow. $NVDA is possibly in the run for $AMBA https://t.co/FbMbGEkQXl

$NBIS $NVDA GOOD GOD. ONE. GIGAWATT. Of Vera Rubin. Next year. Nebius is guiding to a $7-9B run rate by the end of 2026, and they just casually dropped that a gigawatt of the most powerful chips on the planet is coming next, with customers…

BofA: SpaceX management is "highly confident" of adding 10GW in CY27, mostly Vera Rubin, would be surprised below 13GW, and sees "non-zero" chance of more than 20GW. That implies $400 billion to $800 billion of potential Vera Rubin spend…

GTC Berlin After Hours: Keynote Reactions & What’s Next for Developers https://t.co/hbTmwsEGSb

$NVDA $NBIS $IREN $CRWV The world is not ready for the speed of the buildout ahead of us. NVIDIA just unveiled DSX POD at GTC Berlin: complete Vera Rubin AI factories, built and tested inside containers before they ever leave the factory…

Multiple outlets: Nvidia wants next-gen custom HBM base dies on TSMC 3nm, and Samsung may use rival TSMC lines for that die. Unconfirmed. Memory makers keep the stack; the logic die’s foundry choice is becoming a Nvidia-spec decision.…

Jensen continues to march ahead of competitors by locking down the entire supply chain. This is why they can enjoy the pricing power they have. $NVDA https://t.co/ynYeNfsS9f

Nomura: global data center power usage by application $NVDA $AMD $VRT $GOOGL $MSFT $META $AMZN

The memory supercycle is about to get A LOT bigger, Every single new wave of AI is creating an entirely NEW layer of demand for memory, and we're still in the EARLY stages of this cycle. HBM capacity per GPU has already increased 3.6X from…

This.... https://t.co/9viBtstl6u

NVIDIA has released Flora, a set of AI models that turns human and robot videos into training annotations. It identifies individual actions, describes what’s being attempted and scores how well it went, helping researchers turn raw footage…

Cisco and NVIDIA are jointly drafting a new pluggable optics form factor, USFP, built for 400G SerDes lanes rather than today's 200G. Four lanes then carry 1.6Tbps, 16 lanes 6.4T, 32 lanes 12.8T. QSFP and OSFP were not enough. The…

Robotics engineers can now use coding agents to run simulation and training workflows. [📍GitHub - Open Source] @nebiusai just announced an update to its open-source Physical AI Workbench, developed in collaboration with NVIDIA. Anyone who…

A new humanoid platform for robotics research is coming out of Germany. Agile Robots has introduced Agile ONE SN with NVIDIA, combining Jetson Thor, Isaac GR00T and two WUJI Hand 2s. Production is planned to begin this year, with up to…

WHAT 1,730 CRYSTAL PULLERS TELL US ABOUT MUSK’S 20 GW PLAN: Musk says SpaceX has a “highly proprietary” plan to add 20 GW of power by end-2027, beyond just gas turbines. Solar is a part of it and things are moving. China shipped 1,730…

ABF capacity is a major issue for AI buildout. China currently has 4 major replacement option: 华正's CBF, 生益's SIF, 宏昌's GBF & 纽菲斯's NBF Theoretically, Domestic capacity will surpass China's demand by 2028, but key is qualification process…

NVIDIA announces new Isaac tools for agents as Agile Robots unveils Agile ONE SN, a reference humanoid for robotics research. 🦾 Built on NVIDIA Isaac GR00T and powered by Jetson Thor, Agile ONE SN gives researchers and developers an…

NVIDIA already has TensorRT-LLM. It also has Dynamo. This week it shipped a third local inference server anyway, and the part worth sitting with is why. Flare runs on Rust, not Python, and it's built narrow: just the GPUs with the…

🇺🇸 Seeed Studio #PhysicalAI stack is now in stock at @microcenter stores - Columbus, Santa Clara, Cambridge, Dallas, and Houston! Pick up @NVIDIARobotics Jetson-powered reComputer edge AI systems and open-source robotic arms: #reBot Arm…

SpaceX alone could be buying more Nvidia GPUs than most countries. On the earnings call, Elon Musk said SpaceX is on track to bring roughly 200,000 $NVDA GB300 GPUs online every month for the rest of this year. That is about 600,000 GPUs…

You can train Asimov 1 using our open-source RL training framework, built on @NVIDIAAI's Isaac Lab. https://t.co/UR36iZdlKD https://t.co/1f0FtTQ5CV

DeepSeek just gave away the engine it uses to run its own models. The Chinese lab open sourced DeepInfer, the inference framework behind DeepSeek V4.1 Flash, under the free Apache 2.0 license. It runs on $NVDA Blackwell chips and on…

Morgan Stanley: We believe low-volume production of the Nvidia VR200 rack has already begun, with the LPX rack scheduled to ramp late this year or early next year, followed by the Vera standalone CPU rack ramping in late 1Q27 or early…

Morgan Stanley sees the cost per new MW peaking at $51.8 million in FY27, before sliding to $47 million in FY29E. $MSFT $NVDA $AMZN $GOOGL $META https://t.co/IC6zlAbmhs https://t.co/8zXgV111pw

Nvidia is now writing checks to keep its own customers alive. According to The Information, $NVDA is in talks to invest hundreds of millions of dollars in Firmus, the Australian AI cloud provider that just pulled its IPO. Firmus was trying…

Wild. Why we can't get cheap NVIDIA cards. https://t.co/frFkm8HW3A

Fal and Fireworks AI just put out a number worth sitting with: 961.6 tokens per second, one request, on a 27B model, four Nvidia B300s doing the work. That's the early-launch benchmark for EAGLE-4, the fourth generation of a speculative…

Millions of hours of human experience data are being collected. But how do we turn all that raw footage into something models can actually learn from? NVIDIA just open-sourced Flora, a gold-mining tool for Physical AI. It automatically…

The industry believes SK Hynix' HBM4 shipments fell short of initial expectations in the third quarter. According to the industry on the 21st, SK hynix's third-quarter operating profit consensus (the average forecast by brokerages) is…

Who's ready? 👀 #NVIDIAGTC https://t.co/eQFuVJpbRL

Morgan Stanley: We believe AWS's planned incremental GPU purchases will skew to Vera Rubin and Rubin Ultra, we see these extra chips driving an extra ~2 GW /~1 GW of compute capacity in '27/'28 vs. our prior estimates. $AMZN $NVDA

Every hyperscaler racing to dent Nvidia's lead just converged on the same move within a day of each other. Google open sourced TorchTPU at the PyTorch Conference in San Jose this week, co-built with Meta's PyTorch team, letting existing…

$MU $SNDK $SKHY $NBIS Nomura’s memory team just put out a really interesting note on the AI scare, the numbers underneath are some of the strongest I’ve seen in this drawdown. Nomura sees AI going from 10GW of data center power in 2025 to…

Lux family co @modal x @nvidia new Vera CPU... https://t.co/pH6lxp1Y4A

A general-purpose AI model just outperformed specialized robot policies on a manipulation benchmark. Without additional robotics training. Researchers at NVIDIA have published Astranomy, a study looking at how well GPT-6 Astra performs…

“Lack of electricity” is an oversimplification. Everything is hard about bringing massive compute online at extremely high speed. If there was twice as much time, this would be relatively easy, but there is not. https://t.co/rmuECPbCrN

Bernstein on $LITE: Recent drawdowns are due to softer AI lab revenues, $CRWV financing, and macro overhangs. Not weaker infra demand. Bernstein say that the real debate is whether 2028 CPO bandwidth math holds, with bandwidth scale-up…

Expert Takeaways: Data Centers 1. Open-Source Models Closing the Gap > Open-source models are catching up to state-of-the-art frontier models faster than expected, driven by lower costs and improved performance. > Starting in Q2, customers…

Model weights are free. Running them at scale cheaply is not, and that's the part DeepSeek open sourced for everyone this week. DeepInfer is the inference engine serving DeepSeek V4.1 Flash in production right now, and it's fully open,…

How DeepSeek Is Shaping Huawei’s AI Chips in a Growing Challenge to Nvidia https://t.co/14bK0k2Pse

JPMorgan just dropped an ABSOLUTELY INSANE forecast for AI compute. According to JPMorgan, global AI compute demand is expected to reach roughly 530 GW by 2031, while available supply is projected to reach just 250 GW. That leaves us with…

Ropedia just released W0, an open world model trained on ~10,000 hours of egocentric human video. One set of weights predicts first-person views for humans, robot arms, and humanoids (Unitree G1, Fourier GR-1) using a shared camera +…

This is how you get rich ⚡️ You buy the macro dip in companies that are structural to building the future, with great management and a killer product roadmap. This is the time to add, into fear, into bad news, into uncertainty, not to…

This is absolutely INSANE and another example of just how EXTREME demand for AI compute has become. According to Reuters, China's three largest technology companies, ByteDance, $BABA, and Tencent, have reportedly placed more than $35B in…

OpenAI’s officials “were with us on this and now their position is a bit more fuzzy,” said one of the congressional aides granted anonymity. “Silence is doing Nvidia’s work for it.” https://t.co/RwyMk42lnw

Nvidia is expected to retain the largest market share in the AI accelerator market, despite slipping from 81.7% in 2025 to 68.6% in 2030E. Nvidia should remain the leader. $NVDA https://t.co/2nD7AS5EH1

ByteDance, Alibaba, Tencent place over $35 billion worth of Nvidia chip orders US-based chipmaker Nvidia has told customers and partners it cannot guarantee that it can fill the ‌orders because demand from Chinese technology companies far…

Nomura: Sumitomo Electric is on track to secure Nvidia's qualification for ultra-high-power CPO lasers in FY27 before scaling in 2H28.

Startup doubleAI is launching an AI model that it says can alleviate a desperate problem in Silicon Valley: the cost of compute. The company, co-founded by Mobileye CEO Amnon Shashua, says it slashed one early customer’s compute costs by…

Five days left on this one. 🏆 The @nebiusai Physical AI Awards, in partnership with @NVIDIARobotics, are open globally and close 25 October. I’m on the judging panel this year. Five categories, which double as a decent map of how the…

AWS keeps pitching Trainium as a plug-and-play Nvidia alternative, and the actual migration story is more honest than the marketing. PyTorch and JAX code mostly ports to the Neuron SDK with modest changes, not zero changes. Custom CUDA…

NVIDIA、メルセデス・ベンツと自動運転で提携拡大-https://t.co/7W7ZV2mpnU NVIDIAとMercedes-Benzは、運転支援・自動運転システムの共同開発を拡大。AIや車載コンピューティング技術を活用し、将来のレベル4自動運転に加え、生産現場の効率化や品質向上も目指す。

Business Insider: Amazon > Kalyanaraman of AWS Infrastructure expects Amazon to remain free cash flow negative for "the next couple of years or so." > Jassy likes the results of investments with AWS growth ramping as a result. > Internal…

ALERT🚨: SGLANG JUST MERGED MEGAKERNEL SUPPORT, GETTING CLOSE TO 1000 tok/s/user on the GB300 NVL72 AgentX Harness & increasing interactivity by 1.5x on MI355X too! Great work by the SGLang, RadixArk, NVIDIA, and AMD teams! Best of all,…

Amazon says its AI chips will work with popular coding language, challenging Nvidia - https://t.co/MkyybHW1WY

Generated on two DGX Sparks. Wild what you can create on your desk now. https://t.co/VVz8v4WXpp

Nebius: spot pricing of B200 has increased by 109.4% in the past 7 days. $NBIS $NVDA

The AI compute shortage is getting WORSE and what's happening with NVIDIA's B200 GPUs right now is absolutely INSANE. According to Ornn Exchange, rental prices for NVIDIA's B200 GPUs just reached a new all time high of $8.54 per GPU hour,…

🇨🇳Chinese AI startup @deepseek_ai is working with @Huawei to design custom AI chips for its models in a push to train advanced systems without relying on @Nvidia. Huawei engineers have been stationed at DeepSeek’s offices to incorporate…

$SPY rose 0.48% to $766.91 today, with AI and chip stocks leading a bounce after last week's selloff. Here are the top stories you need to know. 1. Chip stocks bounced back as the Philadelphia Semiconductor Index rose about 2% to snap an…

ALERT🚨: NVIDIA has been working overtime, nights, and weekends after being shocked 🤯 by last week’s TPUv7 FP8 submission showing better performance per dollar. Unfortunately for NVIDIA, its first GB300 FP4 counter-submission is still…

We're about to witness the BIGGEST infrastructure spending boom in human history. According to Morgan Stanley, hyperscalers and neoclouds are projected to spend a staggering $1.58 TRILLION in 2027 and $1.81 TRILLION in 2028 when operating…