Tom Gibbs and David Hall, NVIDIA Corporation
Explore how the latest breakthroughs in AI are being applied to problems in climate, weather, satellite remote sensing, material science, biological systems, high energy physics, and space. Deep learning may be viewed as a new method for building GPU-accelerated scientific software that was previously far beyond our reach. We will discuss some of the cutting-edge breakthroughs that have been achieved with these techniques, and examine recent progress addressing major science specific challenges such as interpretability, uncertainty quantification, and enforcing physical constraints. Then we will explore the latest AI trends and discuss the scientific opportunities they present.
