The new large-scale and production cluster packs over 4600 GPUs based on NVIDIA's new GB300 NVL72 architecture, connected through next-gen InfiniBand interconnect fabric. The new deployment allows ...

Quick ReadFrank Bruno argues Nvidia's training-optimized GPU architecture risks the same fate Intel faced when smartphones ...

Nvidia's new co-packaged optics switch offers 3.5x power efficiency and 10x resiliency. Lambda, CoreWeave, Meta, Microsoft, ...

Decentralized artificial intelligence has a lot of ground to make up if it wants to compete with centralized AI providers such as OpenAI Group PBC and Google Cloud, but it’s getting a big boost today ...

Nvidia has invested $6.5B+ in photonics since March, targeting Coherent, Lumentum, Marvell, Corning, and Ayar Labs to replace copper with light in AI data centres.

Uber accelerates incident detection, DCAI speeds up AI training and cluster efficiency, and Nebius improves MTBF in large-scale distributed AI training-all powered by Clockwork's first-of-a-kind ...

Google will pay SpaceX $920 million per month to access capacity to GPUs -- and that's great news for Nvidia.

The AI boom has already burned through one shortage after another. First it was GPUs. Then high-bandwidth memory. Then power infrastructure, transformers, cooling systems, and even land near major ...

Marvell represents the missing nervous system that connects isolated GPU clusters into living, breathing AI infrastructure. The result is a vertically integrated platform no chip rival is replicating ...