NVIDIAEventDevelopersNovember 16, 2026

NVIDIA Vera Rubin debuts, sweeping MLPerf Training v6.1

Read original source →blogs.nvidia.com

Vera Rubin NVL72 delivered up to 2x more performance per GPU than GB300 NVL72 and over 10x throughput per kilowatt in its first MLPerf Endpoints submission. On DeepSeek-V3 it scaled from 256 to 1,024 GPUs at ~97% efficiency, cutting training from 16 minutes to about four.

3 sources

More stories today

Open the live feed