Nvidia Promotes CPUs For Data-Center AI Workloads

Nvidia CEO Jensen Huang said on the company's fourth-quarter earnings call that Nvidia is pushing its data-center CPUs, Grace and Vera, alongside GPUs as AI shifts toward inference and agentic workloads. The company highlighted NVL72 servers (36 CPUs, 72 GPUs) and announced a deal with Meta to use Nvidia CPUs standalone. Nvidia plans to share more CPU details at its developer conference next month.
Key Points
- 1Highlights Nvidia's Grace and Vera CPUs and NVL72 server with 36 CPUs and 72 GPUs
- 2States CPUs better suit agentic, data-driven inference workloads and deployment-phase cost efficiency
- 3Impacts practitioners: expect more CPU-only deployments, vendor competition, and data-center architecture shifts
Scoring Rationale
Official Nvidia CPU push and Meta deal raise industry relevance; limited novelty and mainly strategic, not groundbreaking technical change.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems