NVIDIA Vera Rubin debuts with leading MLPerf performance
NVIDIA
· November 16, 2026
· ✓ verified
NVIDIA has announced that its Vera Rubin platform debuted with leading performance in MLPerf benchmarks across training and agentic inference.
- MLPerf Training v6.1: NVIDIA says it swept every benchmark, and Vera Rubin NVL72 delivered up to 2x more performance per GPU than GB300 NVL72 in its debut.
- MLPerf Endpoints v1.0: In a provisional Kimi K3 result, Vera Rubin NVL72 delivered more than 10x throughput per watt versus GB300 NVL72 at the same response speed; NVIDIA also cited 97% efficiency on DeepSeek-V3 when scaling from 256 to 1,024 GPUs.
- The article references results released today and mentions partner submissions from 20 organizations, plus independent benchmark performance on SemiAnalysis AgentX and Artificial Analysis AgentPerf.