Nvidia's H200 chips have landed in China — in barely traceable numbers
Mira Murati's Thinking Machines Lab releases Inkling, a 975B-param MoE model that uses a fraction of tokens compared to Nvidia's Nemotron.
NVIDIA
Most RecentUS Allows 10 Chinese Firms to Buy H200 Chips, But Nothing Ships Yet
Mira Murati's Thinking Machines Lab releases Inkling, a 975B-param MoE model that uses a fraction of tokens compared to Nvidia's Nemotron.
Nvidia Finally Makes a Card You Can Afford – a Trading Card
Nvidia's Blackwell DGX B300 servers are now going for $1.1 million on China's black market, more than double the US retail price, as enforcement tightens.
Nvidia AI Chip Prices Double on China's Black Market
Nvidia's Blackwell DGX B300 servers are now going for $1.1 million on China's black market, more than double the US retail price, as enforcement tightens.
China May Let Alibaba Buy Nvidia H200s — But Only for Training
Cognition's new SWE-1.7 coding AI model, based on Kimi K2.7 Code and running on Cerebras chips, outputs 1,000 tokens per second and can handle tasks lasting up to 6 hours.
Nvidia unveils Vera, a CPU designed for AI agents
Cognition's new SWE-1.7 coding AI model, based on Kimi K2.7 Code and running on Cerebras chips, outputs 1,000 tokens per second and can handle tasks lasting up to 6 hours.
MIRA is a 5B-parameter world model that runs Rocket League 2v2 in real time
Li Bojie denies 'fleeing with funds,' says ABCDE Capital only wired $500K of a promised $1.5M, leaving his startup struggling.
Nvidia CEO Jensen Huang Is Auctioning His Tom Ford Leather Jacket
Alibaba Cloud's Qoder Enterprise Edition can parse codebases with up to 100,000 files, offering team collaboration, security layers, and flexible pricing.
Nvidia's TwoTower lets two 30B models write in parallel for 2.4x speed
Google’s TabFM can predict customer churn or fraud from structured data without any model training or feature engineering — just feed it a table.
Nvidia's AI Compute Partnership: guaranteed GPU sales for a revenue cut
Google’s TabFM can predict customer churn or fraud from structured data without any model training or feature engineering — just feed it a table.