Nvidia's board reportedly pushed out CEO Jensen Huang mainly over his repeated lying, not just the affair allegations. Sources say he lied about his relationships and what he told the board over about a year.
A column argues that the bigger danger from AI is not rogue robots but powerful tech people. It uses the example of cricket, where a US-based team reportedly uses AI for player picks and in-match decisions.
A report from The New York Times says Alice Huang, at the center of the Nvidia chip scandal, has many ties through family and associates to Chinese state defense contractors. The companies linked to her received about $3.6 billion in Chinese government money, according to a research firm's review. It remains unclear whether Nvidia chips were actually used for military projects.
Morris Chang, the founder of TSMC, built the foundry model that made chip design companies possible. TSMC now dominates contract chipmaking, and its advanced packaging technology CoWoS has become a key bottleneck for AI chip supply.
Nvidia reported healthy growth in its latest earnings report, even as it deals with fallout from ousting CEO Jensen Huang. The chipmaker also said it will stop selling its H200 chips in China and rejected calls to bring Huang back. TechCrunch's Equity podcast discussed what this means for Nvidia and the wider AI industry.
Nvidia is reportedly nearing a deal to buy AI startup Inception for at least $500 million. The startup builds language models that refine a full answer step by step instead of one word at a time. The deal would also include bonuses for its 30-plus employees.
A podcast episode looks at a string of tech failures in one week, including a days-long Outlook outage, a nationwide ground stop, bank disruptions, and iCloud problems. The hosts argue most of the issues were milder than rumors claimed, but they may still point to bigger risks as AI spreads. The episode also covers the Jensen Huang and Nvidia scandal over alleged chip smuggling to China.
The New York Times podcast Hard Fork talks with reporter Joseph Bernstein about the new Grand Theft Auto game, which he got to try in early access. The hosts discuss what the game says about the current tech moment and whether AI might mean we never see a game like this again. They also cover the fallout from Jensen Huang's departure from Nvidia and the news that Anthropic went public on Wall Street as a $2 trillion company.
A report says Nvidia co-founder Jensen Huang has belittled employees for decades. One past case ended in a confidential settlement with an executive who said he terrorized staff.
A weekly roundup covers several chip industry news items, including Intel's first outside customer for its 18A-P chip process, Siemens buying Breker, and new export controls on China.
Samsung Electronics and TSMC have stopped shipping some chips to certain Chinese customers while they wait for US export permits. The move follows new US rules that limit how Chinese firms get advanced chips and have them made overseas. Chinese companies' plans to build their own AI chips could be delayed as a result.
Jensen Huang, NVIDIA's co-founder who long championed robotics, has stepped down as the company's CEO. CFO Colette Kress is serving as interim CEO while NVIDIA searches for a permanent successor.
Microsoft CEO Satya Nadella said he still has confidence in Nvidia's team even after Jensen Huang stepped down as CEO. He said the team can keep the company's performance going. Nadella also said he has kept talking with the same Nvidia people as before.
A US congressional commission reported that China's military has tried to get Nvidia chips by going around US export controls. The report says China's own AI chips still lag behind American ones, which is why it is pursuing this route.
Lawmakers released confidential records showing the U.S. Commerce Department approved sales of more than 800,000 advanced AI chips to Chinese companies through May 11, including some firms tied to the Chinese military. Nvidia also applied to sell chips to a company linked to Megaspeed and a figure who U.S. intelligence says was a Chinese asset.
At Microsoft Ignite on November 19, 2026, NVIDIA announced it is expanding its AI partnership with Microsoft. The plan covers Vera Rubin-based cloud machines on Azure, wider support for NVIDIA's Nemotron open models, and deployment options for on-premises systems, industrial edge devices, and Windows cloud PCs.
After Jensen Huang resigned as Nvidia CEO following reports of several extramarital affairs, some food vendors in Taiwan who had hosted him say they still back him. Shaved-ice seller Tu Chang-chun says he has no plans to take down photos of Huang at his stall, and a tofu pudding shop also keeps its photos up.
NVIDIA explains how RL training slows down during rollouts and how its Dynamo serving layer tackles this. Dynamo routes generation requests to the right workers, speeds up weight updates, and helps recover failed workers. The goal is to finish training iterations faster and waste less GPU capacity.
In this short video, Langdock co-CEO Judith Dada talks with NVIDIA's Chris Alexiuk over a drink. She argues that AI's biggest impact may show up in better services, better products, and healthier, more affordable lives.
NVIDIA and Microsoft announced wider access to their AI computing systems on Azure, plus new tools for building and running AI models. The news also covers AI running on local devices, including Windows PCs. Microsoft said it uses NVIDIA technology in its own models and search services.
This video is a conversation between Langdock Co-CEO Judith Dada and NVIDIA's Chris Alexiuk about Europe's AI future. They cover building computing infrastructure and bringing AI benefits into everyday life.
This short clip from an NVIDIA GTC panel compares enterprise AI security to childproofing a home. The speaker says companies should secure their systems before letting AI run freely.
NVIDIA's DAQIRI library streams sensor data straight into GPU memory so it can be processed while it is being collected. Berkeley Lab's Molecular Foundry used this with a new electron microscope camera that produces over 500 Gbps of data.
NVIDIA published a guide on writing tool-calling parsers in Rust with AI agents. A parser sorts a model's raw output into visible text, reasoning, and tool calls that an app can use.
Amazon AWS Machine Learning blog · Nov 20, 01:13 KST
AWS published a guide on running NVIDIA Run:ai GPU Fractioning on Amazon EKS so several workloads can share one GPU. In a test on one A10G GPU, two Pods each got 6 GiB within an 8 GiB limit, and an allocation above that limit was rejected. The method adds enforced GPU-memory limits and per-Pod metrics to software-based GPU sharing.
NVIDIA's GeForce NOW cloud gaming service is running a Black Friday sale, with 30% off the first six months of Performance or Ultimate memberships through November 30. It also launched on Amazon Fire TV Stick devices running Vega OS, with a 90% off first-month offer through Amazon.
The New York Times has rounded up highlights from its DealBook Summit. Topics include AI, fears of a bubble, the fallout around Nvidia, prediction markets, and politics.
Elon Musk said replacing Nvidia CEO Jensen Huang would be a bad decision, comparing it to benching Michael Jordan from a winning team. He also said he regrets spending at least $474 million on the US midterm elections and that a SpaceX-Tesla merger is not yet decided.
Nvidia reported third-quarter revenue of $115.8 billion, more than double the same period last year. The results came under interim CEO Colette Kress, who took over after co-founder Jensen Huang resigned. Demand from cloud providers and tech giants for its AI processors stayed strong.
Alibaba, a major Chinese internet company, used its AI computing technology to help a lab linked to China's military. The report says Nvidia chips ended up within reach of China's military.