Digital Event Horizon
NVIDIA has made a groundbreaking achievement in the field of artificial intelligence (AI), sweeping the MLPerf Training and Endpoints benchmarks with its new platform, Vera Rubin. With leading performance across training, post-training, and agentic inference, Vera Rubin delivers more than 10x generational gains in throughput per watt, making it an ideal choice for organizations looking to accelerate their AI research and development efforts.
NVIDIA's Vera Rubin platform sweeps the MLPerf Training and Endpoints benchmarks, achieving leading performance across training, post-training, and agentic inference.Vera Rubin delivers more than 10x generational gains in throughput per watt, outperforming competitors in every benchmark.The platform trains AI models to a higher level of accuracy, with some benchmarks demonstrating up to 2x more performance per GPU than competitors.Vera Rubin excels in agentic inference, delivering more than 10x the throughput per watt of its competitors.The platform's success is attributed to its use of NVIDIA's CUDA-powered software stack and open source libraries.Vera Rubin delivers leading performance on independent benchmarks, including SemiAnalysis AgentX and Artificial Analysis AgentPerf.The platform's capabilities have significant implications for the field of AI, enabling faster training times and increased performance.
NVIDIA has made a groundbreaking achievement in the field of artificial intelligence (AI), sweeping the MLPerf Training and Endpoints benchmarks with its new platform, Vera Rubin. This development marks a significant milestone in the company's efforts to advance the state-of-the-art in AI research and development.
According to the results released on November 16, 2026, Vera Rubin delivers leading performance across training, post-training, and agentic inference, outperforming its competitors in every benchmark. The platform's performance is measured in terms of throughput per watt, with Vera Rubin achieving more than 10x generational gains in this metric.
The MLPerf Training benchmark assesses a system's ability to train AI models to a target level of accuracy. In this round, Vera Rubin's performance was evaluated on a range of AI models, including large language models, image generation, recommender systems, and reinforcement learning post-training. The platform's results show that it can train these models to a higher level of accuracy than its competitors, with some benchmarks demonstrating up to 2x more performance per GPU than the highly optimized NVIDIA GB300 NVL72 system.
In addition to its performance in training, Vera Rubin also excels in agentic inference, which is the ability of an AI system to reason, call tools, and check its work to complete a task. The platform's performance in this area is measured in terms of throughput per watt, with Vera Rubin delivering more than 10x the throughput per watt of its competitors.
The success of Vera Rubin can be attributed to its use of NVIDIA's CUDA-powered software stack, which provides a programmable platform for AI research and development. The platform's open source libraries and frameworks allow a vast ecosystem of partners to innovate and optimize AI training performance.
NVIDIA partners participated extensively in the MLPerf Training benchmark, with submissions from 20 organizations, including ASUS, Azure, Cisco, CoreWeave, and many others. These submissions demonstrate how NVIDIA's platform can be used to advance the state-of-the-art in AI research and development.
The lead extended beyond the MLPerf Training benchmark, with Vera Rubin also delivering leading performance on independent benchmarks, including SemiAnalysis AgentX and Artificial Analysis AgentPerf. These results show that the platform can productively run a range of workloads, from pretraining and post-training to agentic inference, and workloads beyond AI, such as high-performance computing.
The success of Vera Rubin has significant implications for the field of AI, with faster training times and increased performance enabling smarter agents to reach users sooner, drug discovery models to keep pace with new lab data, and fraud-detection models to adapt to new threats faster.
In conclusion, NVIDIA's breakthrough in AI research and development with Vera Rubin marks a significant milestone in the company's efforts to advance the state-of-the-art in AI. The platform's performance and capabilities make it an ideal choice for organizations looking to accelerate their AI research and development efforts.
Related Information:
https://www.digitaleventhorizon.com/articles/NVIDIA-Sweeps-MLPerf-Training-and-Endpoints-with-Leading-Performance-A-Breakthrough-in-Artificial-Intelligence-deh.shtml
https://blogs.nvidia.com/blog/mlperf-training-endpoints-vera-rubin/
Published: Mon Nov 16 16:14:11 2026 by llama3.2 3B Q4_K_M