TL;DR: NVIDIA will ship its new B30 AI GPU in Q4 2024, offering 10-20% lower performance than the H20 but at 30-40% reduced cost. With eased US export restrictions, NVIDIA expects 400,000 H20 units ...

Two tiny boxes, 128 GB apiece – but very different strengths Hands On Most GenAI models are trained and run in massive datacenter clusters, but the ability to build, test, and prototype AI systems ...

The new NVIDIA RTX PRO 5000 72GB Blackwell GPU is here, accelerating local AI with 50% more VRAM than the existing RTX PRO ...

As AI becomes more like a recurring utility expense, IT decision-makers need to keep an eye on enterprise spending. The costs of GPU use in data centers could track with overall costs for AI.

Correction, Sept. 24: This article has been edited to correct details about the flow of data through a system containing both a Rubin AI and Rubin CPX GPUs. At the InfraAI Global Summit’25, Nvidia ...

Much attention has been focused in the news on the useful life of graphics processing units, the dominant chips for artificial intelligence. Though the pervasive narrative suggests GPUs have a short ...

Nvidia continues to maintain a firm grasp on the entire GPU market, thanks to decades of innovation and some truly brilliant product launches. Those times are seemingly reserved for the dustbin of ...

With an AMD Ryzen AI Max 395, Radeon 8060S, and 128 GB RAM, the GMKtec EVO-X2 is equally ideal for local AI workflows and ...