NVIDIA's Ian Buck opened the company's week at SC26 in Chicago with new tools for AI agents that can plan experiments, run code, and check results on researchers' own computers. The announcements covered a science model package, new processors for supercomputers, and a quantum computing partnership with IBM and Fermilab.
Nvidia reports quarterly earnings with Wall Street on edge after CEO Jensen Huang resigned last week. Colette Kress is now interim CEO, and analysts want to know about China sales, a permanent successor, and whether the chip business holds up. Nvidia's stock has fallen 27 percent since late October.
The US government has confirmed it is investigating Megaspeed, a Singapore-based computing company. It suspects the firm helped send Nvidia AI chips to China in violation of US export controls. Megaspeed says it has started its own internal investigation.
A New York Times opinion piece argues that the Nvidia scandal shows the Trump administration ignored national security while claiming to win the AI race with China. The writer says the administration loosened AI rules, often citing that race. The article text was partly cut off, so the summary covers only the visible part.
Researchers from Georgia Tech, NVIDIA Research, and Stanford University published a paper called SlipStream. It is a way to make High-Bandwidth Flash (HBF) work well as memory for GPU systems running large language models. Flash memory can hold much more data than DRAM-based memory, but it responds much more slowly.
NVIDIA unveiled the DGX Spark HPC Developer Kit at the SC26 supercomputing conference, aimed at researchers and developers. It links four DGX Spark systems directly so they work as a small cluster on a desk. The kit is on display at the event and is available for sale.
Nvidia's interim CEO Colette Kress will present the company's third-quarter results on Tuesday, the first time without founder Jensen Huang, who resigned last week. Investors and lawmakers want answers about Huang's future, reports that Nvidia chips were smuggled to China through a company linked to him, and the company's AI deals.
NVIDIA's operations leader Deb Shoquist runs the supply chain and factories behind its AI products, working with a global network of manufacturing partners. The article describes how those partnerships were built and how they keep up with each new generation of technology.
Nvidia's board is reportedly not considering bringing back Jensen Huang as CEO. Its focus is on an investigation into his conduct. Huang resigned last week, and many people, including tech leaders, have asked for him to return.
OpenAI CEO Sam Altman said the AI industry will be fine even without Nvidia CEO Jensen Huang, while praising Huang's big role in the AI boom. Huang resigned last week after admitting an affair with an alleged Chinese spy. Altman also said he is not worried about Nvidia's large financial guarantees to OpenAI.
NVIDIA marks 20 years of CUDA, its platform that lets developers use the GPU's parallel power for many kinds of work. The video celebrates the people who built and used it. It premiered at NVIDIA's special address at the SC26 conference in Chicago.
Donald Trump said he would not mind seeing Jensen Huang return as Nvidia's CEO, backing a tech industry campaign for his reinstatement. Huang resigned last week after reports of extramarital affairs, including one with an alleged Chinese spy.
Lambda, a cloud provider, is using NVIDIA Photonics co-packaged optics switches to link NVIDIA GB300 NVL72 racks with more than 10,000 GPUs. The setup aims to cut power use and make the network more reliable for large AI factories.
NVIDIA's new Vera Rubin platform took first place on every benchmark in the MLPerf Training round and posted strong results in a new MLPerf Endpoints test. NVIDIA says the chips use power more efficiently for AI agents, meaning they can do more work per unit of electricity. Results were released in mid-November 2026.
NVIDIA has expanded its CUDA-X libraries, adding tools such as cuPhoton, cuPotential, and cuSpa to help scientists analyze large datasets faster. The company says AI agents can use these tools to speed up research.
NVIDIA's new Vera Rubin NVL72 system finished the latest MLPerf Training round, a standard AI training test, faster than its previous-generation GB300 system on most tests. It set the fastest time on six of seven tests and was up to 2x faster per GPU. On the largest test, a GB300 cluster with 10,240 GPUs posted the fastest time overall.
NVIDIA submitted provisional results to the MLPerf Endpoints v1.0 benchmark for its Vera Rubin NVL72 system. The company says it handles over 10 times more work per kilowatt of power than the GB300 NVL72 at a set responsiveness level. The test covers AI agents that run coding and business workflows.
NVIDIA published a guide on post-training its Nemotron 3.5 Super VL model for scientific work. It uses drug-metabolism prediction as an example, where the model predicts the product a molecule becomes after an enzyme acts on it. The reward comes from experimental data, scored by how similar the predicted molecule is to the real one.
Sam Witteveen tests the HP ZGX Fury, a desktop machine built around NVIDIA hardware, to see which AI models it can run and how fast it runs them. He also shows how people could use it for everyday AI tasks. The video is sponsored by HP and NVIDIA.
Nvidia CEO Jensen Huang resigned last week amid a scandal over his multiple affairs, and Nvidia's shares fell 14 percent. New reports keep surfacing, including alleged chip sales to China and a possible comeback push.