Nvidia this week unveiled new capabilities for Earth-2, a family of AI models that the company describes as the world’s first fully open, accelerated weather AI software stack.
The platform is a collection of models, libraries, and frameworks that Nvidia says makes weather and climate AI more accessible than ever before, with potential users ranging from scientists, startups, and developers to enterprises and governments.
Nvidia said that with access to these pre-trained models and libraries, users can accelerate every step of forecasting, from processing initial observational data to producing two-week global forecasts.
“Historically, weather forecasting has relied on powerful supercomputers running physics-based models,” Nvidia wrote in an article. blog post About the launch. “AI-powered weather forecasting saves significant computational time and cost, allowing more countries, weather enterprises and businesses to run application-specific forecasting systems.”
Specifically, Nvidia has announced the availability of three new models in the toolset; Earth-2 Medium Range, Earth-2 Nowcasting and Earth-2 Global Data Assimilation.
Earth-2 Medium Range is a 15-day, high-accuracy forecast model built on Nvidia’s new Atlas architecture. Designed for medium-range forecasting, the model covers more than 70 weather variables, including temperature, pressure, wind and humidity.
Earth-2 Nowcasting uses a generative AI model called StormScope to deliver kilometer-scale forecasts for local storms and hazardous weather.
Earth-2 global data assimilation, driven by the HELDA model, generates the initial atmospheric conditions necessary for weather prediction. Using GPUs instead of traditional supercomputers, the system can produce global snapshots of current conditions in seconds, Nvidia says improving both the speed and accuracy of forecasts when paired with the Earth-2 model.
These join Nvidia’s existing Earth-2 open weather models CORDIFF and ForecastNet3, which are designed for additional reporting resolution and forecast accuracy, respectively.
Early customers include AI weather startup Brightband as well as national and commercial forecasters such as the Israel Meteorological Service, Taiwan’s Central Meteorological Administration, The Weather Company, and the US National Weather Service.
TotalEnergies, Eni, GCL and Southwest Powerpool, as well as financial risk and intelligence firms AXA and S&P Global Energy, in collaboration with Hitachi, are also early use cases.
“NVIDIA Earth-2 is a major step forward in how advanced weather intelligence can be conducted at scale,” Emmanuel Le Borgne, climate and weather forecasting product manager at TotalEnergies SE, said in a blog post. “Models like Earth-2 Nowcasting are phenomenal for our business because they improve short-term risk awareness and decision making in energy systems where minute and local impacts matter.”
Nvidia said the entire model family will be licensed for commercial and non-commercial use and will be available on GitHub and Hugging Face upon release.
