Chips & Compute5 min reading time

Nvidia Wants to Own Every Chip Inside AI Data Centers

Wired
Read full post
Nvidia unveiled its Vera Rubin chip system, combining CPUs and GPUs to better support AI workloads, aiming to supply complete AI data center hardware. The system offers one CPU per two GPUs and is designed for easier integration, with OpenAI already using it. Nvidia is expanding beyond GPUs to provide CPUs for AI agents, reflecting industry shifts toward complex AI systems.

More on this story


More in Chips & Compute

DOJ Probes Nvidia’s Tie-Up With Groq on Antitrust Concerns

Covered by 2 sources
Chips & Compute5 min read

D-Matrix Connects Raptor XPUs to NVIDIA AI Factories via NVLink Fusion

Covered by 2 sources
Chips & Compute5 min read

Celero raises $275m to bring coherent optics inside AI data centres

The Next Web