Trending storyDeveloping
Vera Rubin NVL72 agent inference benchmark released
1 report1 sourceupdated 12 hours ago
What happened
AI digest
On October 15, 2026, Artificial Analysis released AA-AgentPerf for the Vera Rubin NVL72, measuring its inference performance on agentic coding workloads and comparing it with the GB300 NVL72. Under that workload, the report says Vera Rubin NVL72 leads GB300 NVL72 by up to about 5.7x in throughput per MW. The report covers only this benchmark and that peak throughput advantage; no other workloads or more detailed test results are disclosed.
Written by AI from the coverage · updated 2 hours ago
Coverage
Follow the reports to see the story from different sides.
Oct 16
- AI HOT · ModelsArtificial Analysis 评测 Vera Rubin NVL72，智能体推理每 MW 吞吐领先 GB300 最高约 5.7x
Artificial Analysis 发布 Vera Rubin NVL72 的 AA-AgentPerf 评测，在智能体编码负载中，其每 MW 吞吐相对 GB300 NVL72 最高领先约 5.7x。
Heat over time
Not enough continuous observations to draw a trend yet.