Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

NVIDIA

AI overview

Sign in and the AI will write an overview from our coverage.

Headlines · 1493

  1. Nvidia Halts H200 Sales in China and Rejects Calls to Restore Jensen Huang

    Nvidia said it will stop selling H200 AI chips in China and dismissed calls to bring ousted CEO Jensen Huang back. The company also reported healthy growth in its latest earnings and expressed confidence that good times will continue.

    TechCrunch AI · 🔥 0
  2. Amodei Praises SpaceX Compute Quality

    Anthropic CEO Dario Amodei told CNBC that SpaceX’s compute is “high quality” and that Anthropic is enjoying working with SpaceX, according to Wccftech. His comments temper disputed reports that Anthropic had discussed ending its SpaceX compute deal after the second quarter of 2027. Musk separately said Anthropic was interested in renting another 10 gigawatts of compute from SpaceX.

    Wccftech · 🔥 0
  3. AI Selloff Puts Nasdaq 100 Near Bear Market

    The Nasdaq 100 fell 5.6% on Friday and lost 10.4% for the week, its worst run in more than six years, amid fears that the AI investment boom may be slowing. Nvidia sank 14%, while Anthropic slumped after soaring in its public-market debut. The broad selloff pushed the tech-heavy index to the brink of a bear market.

    Bloomberg Technology · 🔥 3
  4. NVIDIA Targeted by Unverified Ransomware Extortion Threat

    Wccftech reports that the ransomware group The Gentlemen added NVIDIA to its dark-web extortion website on November 18, 2026, and threatened to release allegedly stolen company data on November 28. NVIDIA says it has found no indication that its systems or networks were affected and is investigating a possible security incident involving an external partner. The group has not provided sample files, descriptions of the information it claims to have stolen, or details about the alleged breach. There is also no evidence suggesting that NVIDIA's GeForce products, graphics drivers, or consumer gaming systems were affected.

    Wccftech · 🔥 0
  5. Frontier Models Show Signs of Continual Learning

    A 36Kr analysis cites Anthropic’s engineering report “The components of agent memory,” which found that recent frontier models could select, organize, and verify useful memories in minimally prescribed single-session and multi-session setups. The article also cites Nvidia research on self-summarizing context and GPT-6 Astra results on ARC-AGI-3, using them to argue that continual learning may be emerging across model providers rather than requiring a dedicated memory architecture. These broader conclusions remain interpretive and are based largely on selected studies reported in one secondary article.

    36氪 人工智能 · 🔥 0
  6. 51WORLD Reportedly Files for A-Share IPO Guidance

    A report from TMTPost says digital-twin and physical-AI company 51WORLD completed A-share IPO guidance filing on November 19 and plans to list on the Shenzhen Stock Exchange’s ChiNext board, with Guotai Haitong serving as sponsor. The report says 51WORLD listed in Hong Kong less than 11 months earlier, has remained loss-making for five consecutive years and reported a net loss of RMB182 million for 2025. The filing and proposed listing venue have not been independently confirmed in the supplied report.

    钛媒体 · 🔥 0
  7. Report speculates Jensen Huang may launch a new semiconductor company

    Semafor reporter Rohan Goswami predicted that former NVIDIA CEO Jensen Huang may start a new semiconductor company after a break, according to cnBeta. He also speculated that Huang could raise $50 billion from friends and industry allies for the venture, while emphasizing that no concrete plan, fundraising, or company formation had been announced. Deirdre Bosa separately said Huang resigned rather than being directly fired by NVIDIA's board.

    cnBeta · 🔥 0
  8. Reports say Google is splitting TPU design from direct HBM procurement

    IT之家, citing Chosun Biz, and 36氪 report that Google is separating some TPU design work from HBM procurement. Counterpoint Research, as cited, says Google would design the compute die while MediaTek handles the I/O die, while Broadcom remains involved in developing and packaging some other high-end TPU programs. Bernstein estimated that the new procurement model could save about $50 billion over the TPU product lifecycle; the figure is an analyst projection.

    IT之家 AI · 🔥 0
  9. Phosphen streams PS5 GTA 6 at 60 FPS with Veyra frame generation

    Developer Phosphen announced an experimental PC-side streaming scheme for the PS5 version of GTA 6, using the open-source Veyra tool and neural frame generation to display streamed footage at 60 FPS. Phosphen’s test used about 80 seconds across 14 scenes, with 2,351 distinct frames expanded to 4,700 and average frame rate rising from 29.5 to 59 FPS on an RTX 5080. IT之家 reports that PS5 Remote Play remains limited to 1080p, while post-compression processing cannot restore lost detail and may produce artifacts during fast movement; the PS5 still runs the game at 30 FPS, so input latency is not reduced.

    IT之家 AI · 🔥 0
  10. Nvidia Scandal Fallout Creates Opening for India’s AI Contenders

    Bloomberg Technology reports that geopolitical turmoil surrounding Nvidia’s scandal has created an opening for Indian AI companies. India, long a relative lightweight in the global AI race, is now seen as having an opportunity amid disruption across the industry.

    Bloomberg Technology · 🔥 0
  11. Lenovo CFO Optimistic on AI Server Demand Amid Tech-Stock Volatility

    Lenovo CFO Winston Cheng said the company remains optimistic about demand for its AI servers despite recent volatility in tech stocks. He spoke on Bloomberg: The China Show.

    Bloomberg Technology · 🔥 0
  12. Nvidia Falls for a Sixth Session as Leadership Uncertainty Weighs on the Stock

    Nvidia closed Thursday down 3.3% at $145.98, extending its decline to six consecutive trading sessions and a cumulative 16%, according to cnBeta. Dow Jones Market Data showed the company lost $669 billion in market capitalization during that period. The decline came amid uncertainty over Nvidia’s management and future direction, despite strong quarterly results and raised full-year guidance. Futurum Equities chief market strategist Shay Boloor said there was little evidence that compute demand was slowing and that GPU rental prices were rising.

    cnBeta · 🔥 0
  13. VeloxCon China 2026 to Focus on AI Execution Infrastructure

    A report says VeloxCon China 2026 is scheduled for December 5 at Alibaba’s Xuhui Binjiang campus in Shanghai. The reported agenda includes GPU acceleration, agent infrastructure and unified multi-engine architecture. It also lists experts from Meta, Alibaba Cloud, NVIDIA, IBM, Google, Tencent and Ant, though their participation is not independently confirmed in the supplied report.

    掘金 · 人工智能 · 🔥 0
  14. SpaceX closes $32B Nvidia chip debt deal; lenders eye orbit ban on collateral

    SpaceX closed roughly $32 billion of bond financing under Project Saturn, part of a $43-44 billion package led by Goldman Sachs and Apollo to buy Nvidia GPUs, with the largest confirmed tranche an A3-rated note of about $25 billion at a yield near 8.26%. According to The Financial Times, as relayed by TechNews and cnBeta, lenders sought a contractual bar on launching the collateralized AI chips into space for Musk's orbital data center plans, since chips in orbit would be difficult to repossess and resell if SpaceX defaulted. SpaceX says the financed chips are destined for ground data centers, and it is unclear whether the restriction made it into the signed documents.

    科技新报 · 🔥 13
  15. Moody's Cuts Nvidia Credit Outlook to Stable After CEO Exit

    Moody's Ratings reportedly changed Nvidia's credit outlook from positive to stable, citing governance concerns following Jensen Huang's departure. The agency said the outlook change makes a rating upgrade unlikely over the next 12 to 18 months, while an analyst said Huang's exit created uncertainty around business strategy and financial policy.

    cnBeta · 🔥 0
  16. TMTPost Reports NVIDIA Succession Review After Huang’s Reported Resignation

    A TMTPost commentary says Jensen Huang announced his resignation on November 12 and that NVIDIA began considering successors, including Arm CEO René Haas. It also reports that Jeff Bezos warned against hiring an external CEO, several Wall Street funds asked the board to preserve the possibility of Huang’s return, and employees circulated a petition supporting his reinstatement. The commentary argues that succession is especially difficult because Huang has deeply shaped NVIDIA’s hardware architecture, networking, software ecosystem and supply allocation; it does not independently substantiate these claims.

    钛媒体 · 🔥 0
  17. Nvidia Responds to Gentlemen Ransomware Data Leak Threat

    A supplied report states that Nvidia responded to a data-leak threat from the Gentlemen ransomware organization. It provides no details about the threat, the affected data or systems, the timeline, or Nvidia’s response.

    DoNews · 🔥 0
  18. Nvidia’s Strong Q3 Growth Meets a 52-Week Share-Price Low

    AIbase reports that Nvidia’s third-quarter revenue doubled to $115.8 billion, data-center revenue grew 110%, and the Vera Rubin platform generated nearly $20 billion. The company’s shares nevertheless closed at $145.98 on Thursday, down nearly 40% from their October 6 high and at a 52-week low.

    AIbase AI新闻 · 🔥 0
  19. Claude Tops Singularity-Bench With Five Wins and One Draw

    The Singularity-Bench preprint evaluates six AI-development tasks covering training, data selection, reinforcement-learning code, inference engines, agent runtimes, and GPU cluster scheduling. New智元 and AIbase report that Claude Opus 5.5 with Claude Code was the only combination above the human baseline, with five wins and one draw, while its generated inference engine reduced P90 first-token latency from 12.029 seconds with SGLang to 0.478 seconds on matched hardware. The researchers note that humans still defined each task, so the results do not yet demonstrate end-to-end recursive self-improvement.

    新智元 · 🔥 0
  20. Chip Stocks Fall Amid Nvidia Leadership Rumor and Other Concerns

    TechNews reported that Taiwan's three major institutional investors net sold more than NT$46.1 billion as the Philadelphia Semiconductor Index fell 2.41% and TSMC's ADR dropped 2.90%. The decline was accompanied by an unconfirmed market rumor that Nvidia CEO Jensen Huang had been removed, along with concerns involving Supermicro's China litigation and AMD firmware exports. No official Nvidia clarification on the leadership rumor was available.

    科技新报 · 🔥 0
  21. Ruiyang urges enterprises to turn AI agents into governed business tools

    Ruiyang Information says enterprises should scale AI agents beyond isolated use cases and integrate them with existing ERP, knowledge-management and CRM systems. Its knowledge chief, Li Zhengquan, recommends the AI Technology Foundation’s four-stage maturity model, emphasizing data preparation, permissions, workflow integration, security and human oversight as prerequisites for business value. The article also cites industry research about agent-project failures and data-system barriers, while claims about multi-agent efficiency remain vendor assertions.

    科技新报 · 🔥 0
  22. NVIDIA’s Leadership Moat Still Runs Through Jensen Huang

    TechNews argues that Jensen Huang’s departure exposes NVIDIA’s dependence on the founder’s long-term decisions, especially the sustained investment in CUDA and its developer ecosystem. The commentary also traces NVIDIA’s expansion into AI infrastructure through cuDNN, DGX-1, Tensor Cores and the Mellanox acquisition. It says Huang appears likely to return, but provides no direct confirmation.

    科技新报 · 🔥 0
  23. Nvidia in Advanced Talks to Acquire Inception for at Least $500 Million

    The Information, citing three people familiar with the talks, reported that Nvidia is in advanced talks to acquire diffusion-model startup Inception for at least $500 million, with performance-based earnouts for new employees. Inception develops diffusion large language models that refine full responses step by step instead of generating them token by token, and recently released Mercury 3. The proposed acquisition is intended to bolster Nvidia’s open-source Nemotron models and physical models for robots. Neither Nvidia nor Inception has confirmed the talks.

    The Information · 🔥 0
  24. Robinhood CIO Discusses Anthropic’s Record IPO and Nvidia Trading

    Robinhood Markets CIO Stephanie Guild discussed Anthropic’s record-breaking IPO and what it says about the AI trade in an interview with Bloomberg’s “The Close.” She also said Robinhood customers continued buying Nvidia on a net basis on days the stock fell following recent news, although some sold. Nvidia remains one of the platform’s highest-volume stocks.

    Bloomberg Technology · 🔥 0
  25. Microsoft and NVIDIA expand AI infrastructure from Azure to Windows PCs

    Microsoft and NVIDIA announced broader access to NVIDIA Vera Rubin NVL72 systems in Azure, alongside Nemotron open models and customization capabilities in Microsoft Foundry. They also outlined confidential-computing and power-efficiency work, local deployment options, and Direct Stream for GPU-accelerated Windows 365 Cloud PCs, with a public preview planned for early 2027. NVIDIA reported that Azure scaled to 10,240 GB200 NVL72 GPUs on DeepSeek-V3 in MLPerf Training v6.1.

    NVIDIA Blog · 🔥 0
  26. D-Matrix unveils Raptor AI inference chip with stacked 3D DRAM

    D-Matrix introduced Raptor, a second-generation AI inference chip that stacks 3D DRAM beneath its compute logic. The design builds on the Pavehawk test chip, which is proving out the stacked-memory technology. D-Matrix says the architecture addresses inference memory-bandwidth and energy-efficiency constraints, while also identifying interconnect efficiency and software cohesion as priorities. Srivastava claimed that 3D DRAM can deliver 10 times higher energy efficiency and bandwidth than HBM, but the report provides no independent benchmark results.

    SiliconANGLE AI · 🔥 0
  27. SageMaker HyperPod adds enhanced observability for GPU training

    AWS announced enhanced observability for Amazon SageMaker HyperPod, combining host, GPU, EFA network, and training-job telemetry for workloads orchestrated by Amazon EKS or Slurm. Its diagnostic summary compares GPU utilization, memory, temperature, power, and SM clock to flag deviations, while static rules identify issues such as XID and ECC errors, EFA throughput drops, thermal throttling, and overheating. AWS says the capability is available in all regions where SageMaker HyperPod is available.

    AWS Machine Learning Blog · 🔥 0
  28. Hunterbrook's AMD short report targets circular financing and China ties

    AMD shares fell more than 7% after Hunterbrook Capital disclosed a short position and released a report alleging circular deals, excessive customer commitments and business ties to Chinese companies on a U.S. blacklist. Hunterbrook estimated $28 billion in credit support for partners and $48.8 billion in supplier commitments; it also alleged that AMD issued stock warrants to OpenAI and Meta. AMD disputed the report's characterization of its commercial practices and partnerships and said it complies with U.S. export-control law.

    cnBeta · 🔥 0
  29. Lamont raises US AI bubble probability to 70%

    Acadian Asset Management portfolio manager Owen Lamont estimated in November 2026 that the probability of an AI bubble in US stocks had risen to as high as 70%, reversing his January view that one had not formed. He said his four warning indicators—elevated valuations, bubble belief, stock issuance and new-money inflows—were now present, although net issuance equaled 3.1% of market capitalization, below earlier technology-boom peaks. Lamont said he could not determine whether the issuance shift reflected companies viewing shares as overvalued or a need to finance AI infrastructure, and he did not assert that the market had already peaked.

    虎嗅 AI · 🔥 0
  30. Pete Warden on Local AI and Open-Weight Models

    Pete Warden, co-creator of Moonshine Voice and a longtime deep learning engineer, discusses the shift toward running AI locally instead of in the cloud. The interview covers open-weight models, quantization, LLM execution on laptops and in browsers, low-latency voice interfaces, specialized versus end-to-end models, and hardware approaches that could challenge today's GPU-centric infrastructure.

    Hacker News 首页(100+ 分) · 🔥 0

← PreviousNext →

Experience and discussion from the community

Share my NVIDIA experienceAsk about NVIDIA

Switchyard: An open-source library that helps an AI agent choose which model handles each request.

What problem it solves AI agents need to choose among models while balancing task success, cost, and latency for their workloads. Applications also need flexibility in where routing runs and who controls model clients, credentials, retries, and service operation. Key features Multiple routing algorithms — Routing can be configured with Auto, Task, Execution, or Composite strategies, including mode…

github.comhttps://github.com/NVIDIA-NeMo/Switchyard

vllm: A fast and easy-to-use library for LLM inference and serving.

What problem it solves Serving large language models requires efficient memory management, flexible model execution, and practical integration with existing models and APIs. vLLM addresses these needs with inference and serving capabilities spanning batching, caching, quantization, distributed execution, and multiple hardware platforms. Key features Efficient memory and serving — vLLM manages atte…

github.comhttps://github.com/vllm-project/vllm

skills: Portable instruction sets that teach AI agents how to use NVIDIA software for end-to-end workflows.

What problem it solves AI agents need product-specific instructions to use NVIDIA software reliably across physical AI, robotics, simulation, CUDA-X libraries, RAG, and platform workflows. The catalog centralizes these skills while keeping them maintained in their respective product repositories and synchronized through an automated pipeline. Key features Agent-discoverable skill catalog — The NVI…

github.comhttps://github.com/NVIDIA/skills

SkillEvaluator: An open-source tool that checks AI agent skills and measures their effect on task results.

What problem it solves AI agent skills need to be checked for quality, security, duplication, and their actual effect on agent behavior. Static checks alone cannot establish whether guidance improves task results, so evaluation must also support live comparison with and without a skill. Key features Three-tier evaluation pipeline — Independent tiers validate skill safety and structure, detect over…

github.comhttps://github.com/NVIDIA/SkillEvaluator

llama.cpp: Plain C/C++ implementation for local and cloud LLM and VLM inference.

What problem it solves Running large language and vision-language models can require substantial setup and may depend on the available hardware. The project aims to make inference accessible across a wide range of local and cloud hardware with minimal setup. Key features Minimal-setup inference — The project provides installation options for common platforms, including scripts, Docker, pre-built b…

github.comhttps://github.com/ggml-org/llama.cpp

TensorFold: A native language-model inference engine for serving checkpoints on Apple Silicon and NVIDIA GPUs.

What problem it solves Serving language models requires reading checkpoints, tokenizing requests, and running accelerated kernels while managing a local model server. TensorFold addresses these requirements with a native binary that can run on Apple Silicon and NVIDIA GPUs without requiring Python or MLX for serving. Key features Native Metal and CUDA inference — The Zig-based engine executes mode…

github.comhttps://github.com/ashhart/TensorFold

sglang-omni: A multi-stage serving runtime for omni, speech, and TTS models.

What problem it solves Interactive multimodal and speech systems often require generation to span heterogeneous stages with different compute patterns, dependency structures, and resource needs. Serving these pipelines requires coordinated stage lifecycles, inter-stage transport, model integrations, and a unified API surface. Key features Coordinated multi-stage runtime — Models generation as coor…

github.comhttps://github.com/sgl-project/sglang-omni

polars: An analytical query engine written in Rust for DataFrame and SQL queries.

What problem it solves Data analysis workflows often require expressive query composition, flexible execution models, and interoperability with columnar data. Polars addresses these needs by allowing the DataFrame API and SQL to share the same optimizer and engine, with support for large workloads and custom extensions. Key features Unified DataFrame and SQL Queries — Queries can be written with e…

github.comhttps://github.com/pola-rs/polars

FreeToken: An edge-native MoE serving engine for running frontier-scale open-weight models on personal and consumer hardware.

What problem it solves Running frontier-scale open-weight models locally requires an inference platform that can make efficient use of the heterogeneous hardware available on a personal device. The project addresses this by treating GPUs, CPUs, host memory, and interconnects as a unified, elastic inference platform. Key features Edge-native MoE runtime — The runtime provides MoE serving with bandw…

github.comhttps://github.com/FlashML-org/FreeToken

OpenShell: A safe, private runtime for fleets of autonomous AI agents.

What problem it solves Autonomous AI agents often need to read files, install packages, call APIs, and use credentials, but unrestricted access can expose host data, secrets, and networks. OpenShell provides these capabilities within an isolated runtime where declared policies constrain agent access. Key features Kernel-level runtime enforcement — Agents run in isolated sandboxes where kernel cont…

github.comhttps://github.com/NVIDIA/OpenShell

SkillSpector: Security scanner for AI agent skills.

What problem it solves AI agent skills execute with implicit trust and often receive minimal vetting before installation. SkillSpector helps users assess whether a skill contains vulnerabilities, malicious patterns, or other security risks before installing it. Key features Multi-format scanning — Scan agent skills supplied as Git repositories, URLs, zip files, directories, or individual files. So…

github.comhttps://github.com/NVIDIA/SkillSpector

jasna: Restores mosaics in videos and images through a simple app and command line that run on your GPU.

What problem it solves Mosaics obscure portions of videos and images, making it difficult to view the affected content without changing the rest of the picture. Users need a restoration workflow that can identify selected mosaic regions and process them with GPU-accelerated video handling. Key features Targeted mosaic restoration — Detects mosaics in videos and restores them while leaving the rest…

github.comhttps://github.com/Kruk2/jasna

Model-Optimizer: A library for composing model optimization techniques and exporting optimized quantized checkpoints.

What problem it solves Deep learning models often need to be adapted for efficient deployment across different inference frameworks. Model Optimizer addresses this by providing a unified set of optimization techniques and export workflows for quantized checkpoints. Key features Post-training quantization — Compress models with quantization techniques available for Hugging Face, Megatron-Bridge, Di…

github.comhttps://github.com/NVIDIA/Model-Optimizer

nvke: A self-hosting management plane for creating and operating Kubernetes clusters on NVIDIA GPU infrastructure.

What problem it solves Platform teams need to create and operate isolated Kubernetes clusters for customers on infrastructure they supply and control. They need a management plane that can provision GPU machines, manage cluster lifecycle operations, and establish node identity before issuing join credentials. Key features Independent workload clusters — Create workload clusters through the NVKE AP…

github.comhttps://github.com/dsx-ai-factory/nvke

LocalAI: An open-source AI engine for running models locally across modalities and hardware.

What problem it solves Running AI models locally can require integrating multiple inference engines, model formats, hardware accelerators, and API surfaces. LocalAI addresses this with a small core, on-demand backends, and compatible APIs that work across supported hardware without requiring a GPU. Key features Composable inference backends — Backends are packaged separately and pulled only when a…

github.comhttps://github.com/mudler/LocalAI

TensorRT-LLM: An open-source library for optimizing LLM and Visual Gen inference on NVIDIA GPUs.

What problem it solves LLM and Visual Gen inference requires specialized kernels, efficient runtime orchestration, and flexible deployment across NVIDIA GPU configurations. Developers also need a modular framework that supports customization while integrating with established inference tooling. Key features Optimized inference kernels — Custom kernels accelerate common inference operations, includ…

github.comhttps://github.com/NVIDIA/TensorRT-LLM

cuda-rust: NVIDIA's CUDA platform for Rust, providing host-side runtime crates and GPU kernel programming models in idiomatic Rust.

What problem it solves Rust developers need host-side access to the CUDA runtime and tools for expressing GPU kernels while preserving Rust's ownership and borrowing rules. The platform provides separate programming models for workloads that require direct thread-level control and workloads naturally expressed through tensor operations. Key features Shared CUDA runtime — Host-side crates provide d…

github.comhttps://github.com/NVIDIA/cuda-rust

auto-ontology: An open library that builds a context layer from enterprise metadata and exposes it to agent harnesses.

What problem it solves Agents can retrieve documents and query databases, but unreliable results undermine agentic analytics. Teams need a context layer that records where data lives, what its definitions mean, whether it is usable, how it should be queried, its provenance, and related concepts. Key features Enterprise ontology generation — Generate semantic ontologies from database metadata, infe…

github.comhttps://github.com/NVIDIA/auto-ontology