Fig. 6From: Large scale performance analysis of distributed deep learning frameworks for convolutional neural networksParallel efficiency comparison of PyTorch-DDP on up to 1024 GPUs for different ResNets with DALI data loader (CPU-based) and compressed ImageNet dataset, averaged over three runs. The black line denotes the ideal case. The variance between runs is small (in general \(<5\%\)) and therefore not shownBack to article page