NVIDIA reports that its Vera Rubin platform made its debut in the MLPerf benchmarks, with results released on the stated date. According to the company, the platform delivered leading performance across training, post-training and agentic inference. NVIDIA states it swept every benchmark in MLPerf Training v6.1 and, in its first MLPerf Endpoints submission, posted more than 10x generational gains in throughput per watt for agentic AI, a measure tied to agentic workloads.

Why it matters

MLPerf is a widely referenced set of benchmarks for comparing AI hardware and system performance. A clean sweep of the training benchmarks and large generational efficiency gains for agentic inference, as reported, position the Vera Rubin platform as a point of comparison for training and inference workloads.

Who should care

Teams evaluating hardware for model training, post-training and agentic inference may find these benchmark results relevant when assessing performance and energy efficiency.