This short video from Matt Wolfe is about why it is hard for other companies to catch up with NVIDIA. The description only sets out the topic and does not explain the reasons in detail. Viewers should expect a brief overview of the idea rather than a full analysis.
Scientific computing is entering the era of agentic AI. Bringing simulation, AI, and agents together gives researchers new ways to tackle complex problems and accelerate scientific discovery. Watch Ian Buck, VP of Hyperscale and HPC at NVIDIA, highlight the latest cutting-edge breakthroughs revolutionizing the scientific computing community. https://blogs.nvidia.com/blog/ian-buck-sc26/ https://blogs.nvidia.com/blog/sc26-wrapup
Nvidia's new CEO Colette Kress held her first earnings call after Jensen Huang left the company. She said employees are committed and business is running as usual. The article suggests she may be less free with spending than Huang was.
NVIDIA described a way to build training datasets for physical AI, meaning AI that works with the real world, using its Cosmos Data Factory. A coding agent can run five ready-made blueprints from two prompts, covering data curation, augmentation, and synthetic data generation. The post walks through five use cases.
The AI Podcast is back. Hugging Face’s Thomas Wolf and ServiceNow’s Amanda Grady join us to discuss why building in the open matters for AI safety. Because sharing code means more people can help identify problems and fix them. Listen on your favorite podcast platform or watch on YouTube: https://youtu.be/y9AALMeAeGI?si=zARxzOBGTaG28Jql
NVIDIA's RTX Spark compact desktops from ASUS, Dell and MSI are now on sale. They are Windows PCs meant for running AI models and agents locally, along with image and video generation and model fine-tuning. The article also covers a growing set of creative and development tools that run on RTX Spark.
The US Securities and Exchange Commission (SEC) has opened a civil investigation into Nvidia. It is checking whether the company broke disclosure rules or misled investors about its chip sales in Asia. The probe centers on sales that were tied to China.
Nvidia has stopped shipping its advanced H200 AI chips to China, even though the White House had approved the sales in December. Interim CEO Colette Kress announced the change during an earnings call. She did not explain why, or say whether Nvidia will do other business with China.
Nvidia said its loan guarantees for customers' data centers only start once the customer has secured the financing and equipment setup. If the project is delayed or fails, the guarantee no longer applies. The company's interim CEO also said Nvidia will be more selective about future financing deals.
This NVIDIA podcast clip features Hugging Face co-founder Thomas Wolf and ServiceNow's AI platform security leader Amanda Grady. They discuss keeping AI agents safe when simply telling a model not to do something is not enough.
Nvidia co-founder Jensen Huang resigned after revelations of extramarital affairs, one reportedly involving a Chinese spy. The scandal wiped hundreds of billions of dollars off the company's value and raised questions about its future.
Nvidia said its stakes in private and public companies are now worth $65.5 billion, up from $22 billion in January. Much of the growth came from new investments and higher valuations of companies it already backed. Analysts are unsure whether this pace will last after CEO Jensen Huang steps down.
Nvidia reported third-quarter revenue of $115.8 billion, up 103% from a year earlier and above its own forecast, even after CEO Jensen Huang's departure. Interim CEO Colette Kress also said the company expects fourth-quarter revenue of about $130 billion and raised its full-year outlook.
Amazon Web Services said its SageMaker HyperPod Inference service now supports NVIDIA Dynamo, a tool that splits AI model serving across multiple machines. AWS said the split setup handled about twice the throughput of the best combined setup that met its latency targets in one test.
Amazon AWS Machine Learning blog · Nov 18, 08:33 KST
Amazon SageMaker HyperPod Inference now supports NVIDIA Dynamo, an open source framework that splits LLM serving across machines. The goal is to handle large models and long prompts more efficiently. The post covers a deployment example, benchmarks, and tuning tips.
Nvidia's interim CEO Collette Kress said on the company's earnings call that Nvidia remains committed to open-source AI. She cited fast growth in open model usage and said the planned Hugging Face acquisition is still on track.
Nvidia reported record quarterly revenue of $115.8 billion, more than double last year's level, with data center sales leading the growth. Interim CEO Colette Kress presented the results, and the company expects about $130 billion in revenue next quarter.
NVIDIA published a practical guide for model developers on designing sparse attention, a technique that lets AI models focus on only a small part of a long context. The post explains how the parts of the design affect speed on NVIDIA GPUs. It uses a production FP8 kernel to test design choices.
Jensen Huang, who resigned as Nvidia's CEO last week, said he would be willing to return if the board wants him back. He made the statement after reports about two personal affairs, one involving an alleged Chinese spy. CFO Colette Kress is serving as interim CEO for now.
Nvidia CEO Jensen Huang resigned after reports of several affairs, including one with an executive at Megaspeed, a firm US intelligence has investigated for helping China get chips. The scandal has also exposed a clash inside the Trump administration over how tough to be on China. Nvidia's stock value has already dropped by hundreds of billions of dollars.