Nouveau Recherche PDF, HTML, DOCX et bien d'autres formats.

Prépublication · 2026

Half-precision inference of AI weather prediction models: negligible loss of skill at a fraction of the cost

Stefan Siegert — Monde

Source : arXiv

Résumé

AI weather prediction models are released with 32-bit (fp32) weights and run in fp32 by default, although current GPUs compute much faster in 16-bit formats. We run five publicly available models -- Aurora, Pangu-Weather, FourCastNet, SFNO and FastNet -- on a single 16 GB GPU from 52 initial dates in 2024, in fp32 and with their networks in half precision (fp16), and verify the forecasts against ERA5. The physical state must stay in fp32: converting the whole model rounds it in physical units, which in fp16 overflows and in bf16 discards meaningful information. With the state, its normalisatio…

Actions

Citation