Nvidia Unveils Vera Rubin Platform at GTC 2026

Chief Executive Jensen Huang said the new system delivers 10x higher inference performance per watt and a 90% reduction in inference token costs, while projecting combined Blackwell and Vera Rubin orders above $1 trillion before 2027.

45d ago

Summary

No Summary provided as the original text is short

Terms & Concepts

Inference: The process of running a trained artificial intelligence model to generate outputs such as text, images, or predictions.
Inference performance per watt: A measure of how efficiently hardware performs AI inference relative to its power consumption.
Inference token costs: The expense of generating units of AI model output, often used to gauge operating costs for large language models.