The company’s new Earth-2 platform helps forecasters at every stage, from processing initial observation data to generating 15-day global weather and storm forecasts.
- Nvidia’s first model is Medium Range, which is aimed at weather predictions for up to 15 days in advance, across more than 70 weather variables, including temperature, pressure, wind, and humidity.
- The second model, named Nowcasting, uses generative AI to make country-level forecasts within minutes, offering kilometer‑level detail and a prediction range of zero to six hours.
- Nvidia also announced the Global Data Assimilation model, which is used to produce initial conditions for weather prediction.
Nvidia Corp. (NVDA) on Monday announced the launch of the Earth-2 family of open artificial intelligence (AI) models for weather forecasting.
Nvidia’s new Earth-2 platform helps forecasters at every stage, from processing initial observation data to generating 15-day global weather as well as storm forecasts.
Nvidia shares were down 0.3% in Monday’s opening trade. Retail sentiment on Stocktwits around the company was in the ‘bullish’ territory at the time of writing.
What Is Earth-2 Capable Of?
Detailing the features of the Earth-2 platform, Nvidia announced three models aimed at different needs. The first model is the Medium Range, which provides weather predictions up to 15 days in advance across more than 70 weather variables, including temperature, pressure, wind, and humidity.
“On standard benchmarks, it outperforms leading open models on the most common forecasting variables measured by the industry,” the company said.
The second model, named Nowcasting, uses generative AI to make country-level forecasts of local storms and hazardous weather within minutes, offering kilometer‑level detail and a prediction range of zero to six hours.
Nvidia also announced the Global Data Assimilation model, which is used to produce initial conditions for weather prediction — snapshots of the current atmosphere, including the temperature, wind speed, humidity, and air pressure, at thousands of locations around the globe.
The company says this model can generate these initial conditions within seconds using graphics processing units (GPUs), instead of hours on supercomputers.
Developers Tapping Into Earth-2 Already
Nvidia also said that the Earth-2 platform is already being tapped into by developers across industries.
The company gave examples of The Weather Company and the U.S. National Weather Service, S&P Global Energy, TotalEnergies, Eni, GCL, and Southwest Powerpool in collaboration with Hitachi, among others.
“NVIDIA Earth-2 models give us a 90% reduction in compute time at 2.5-kilometer resolution compared with running a classic numerical weather prediction model without AI on a CPU cluster,” said Amir Givati, director of the Israel Meteorological Service.
NVDA stock is up 0.4% year-to-date and 31% over the past 12 months.
For updates and corrections, email newsroom[at]stocktwits[dot]com.<
