Latest / Science and research

ResearchScienceUS InternationalConfirmed

Google’s NeuralGCM hybrid climate model runs 100,000 times cheaper than a physics model

See how this story connects

Google Research, with the European Centre for Medium-Range Weather Forecasts, published NeuralGCM in Nature on 22 July 2024. The model combines a physics-based atmospheric solver with machine-learned components, and Google reported simulations about 100,000 times less computationally expensive than the high-resolution X-SHiELD model. Code and weights were released on GitHub for non-commercial use.

Why it matters

Climate scientists got an open hybrid physics and machine-learning atmosphere model, built with ECMWF, that Google reports costs about 100,000 times less compute than a high-resolution physics model.

When
22 Jul 2024
ID
WTA-N-2024-0024
Who
Google, ECMWF

Summary by wheresthe.ai, from Google Research and the reporting credited above. Checked against the source on 9 Oct 2026.

Source copy: download the page as saved on 18 Oct 2026 (PDF). Also held by the Internet Archive.

Next stop on this Storyline: Led toDeepMind’s GenCast beats ECMWF’s ensemble forecast on 97.2% of targets

Cite this story

WTA-N-2024-0024: “Google’s NeuralGCM hybrid climate model runs 100,000 times cheaper than a physics model.” wheresthe.ai, 22 Jul 2024. https://wheresthe.ai/d/google-neuralgcm-nature-2024-07/ (source: Google Research).

Spotted an error? Email [email protected].