NVIDIA 的博客指出,即使使用看似相同的 H100、GB200/GB300 NVL72 系统,训练吞吐量仍可能出现明显差异,并给出提升性能的做法。

Two AI computing clusters built from identical NVIDIA H100, GB200 NVL72, or GB300 NVL72 systems can deliver materially different training throughput. We...