NASA has partnered with NVIDIA to build a machine learning model that will work to monitor and forecast air pollution worldwide.
The model feeds air pollution data from the NASA Center for Climate Simulation and houses Rapids data science software libraries and V100 Tensor Core graphic processing units to speed up machine learning algorithms, the company said Thursday.
In addition, the model can connect with full Earth system models. Using various GPU-based systems such as DGX-1 helped minimize the time required to produce detailed real-time air quality maps.
“NASA's global models quickly produce terabytes of data, and what we'd like to do is train the machine learning model on these huge datasets,“ said Christoph Keller, a research scientist at NASA.
The model's air quality forecasts are posted through NASA's website, and the agency intends for the data to be utilized by government agencies and application developers.