MLPerf Training v6.0: Lambda delivers fastest LLM training on NVIDIA GB300 NVL72 and fastest MoE training on NVIDIA HGX B200

In MLPerf Training v6.0, Lambda’s NVIDIA GB300 NVL72 Llama 3.1 8B submission achieved a 18.7% performance improvement and the fastest convergence. Lambda also posted the fastest single-node NVIDIA HGX B200 results for GPT-OSS-20B MoE training.