Exascale Deep Learning for Climate Analytics

TLDR

The study extracts pixel‑level masks of extreme weather patterns using Tiramisu and DeepLabv3+ neural networks. The authors enhance software frameworks, input pipelines, and training algorithms to efficiently scale these networks on the Piz Daint and Summit supercomputers. The Tiramisu network achieves 21.0 PF/s on 5 300 P100 GPUs (79 % efficiency), while DeepLabv3+ reaches 325.8 PF/s on 27 360 V100 GPUs (90.7 % efficiency) and 1.13 EF/s peak / 999 PF/s sustained in FP16.

Abstract

We extract pixel-level masks of extreme weather patterns using variants of Tiramisu and DeepLabv3+ neural networks. We describe improvements to the software frameworks, input pipeline, and the network training algorithms necessary to efficiently scale deep learning on the Piz Daint and Summit systems. The Tiramisu network scales to 5300 P100 GPUs with a sustained throughput of 21.0 PF/s and parallel efficiency of 79.0%. DeepLabv3+ scales up to 27360 V100 GPUs with a sustained throughput of 325.8 PF/s and a parallel efficiency of 90.7% in single precision. By taking advantage of the FP16 Tensor Cores, a half-precision version of the DeepLabv3+ network achieves a peak and sustained throughput of 1.13 EF/s and 999.0 PF/s respectively.