During a keynote address Monday at the Berlin Summit for the Earth Virtualization Engines program, Nvidia founder and CEO Jensen Huang told delegates that AI and faster computing will assist climate researchers in performing the miracles they need to accomplish to make advances in climate research.
According to a post on the Nvidia website, during the summit, held at the Harnack House in Berlin, a popular place for the research and scientific community to gather, Huang delivered the keynote to 180 delegates in attendance at the summit.
“Richard Feynman once said that "what I can't make, I don't understand" and that is the reason why climate modeling is so crucial,” he said in the address, according to the Nvidia website. “Therefore, he continued, "the work you do is crucially important to policymakers, researchers, and the industry."
The summit, according to the post on the Nvidia website, draws attendees from across the globe to promote work using AI and high-speed computers to predict climate shifts.
Moreover, according to the website, Huang told delegates three miracles are needed to ensure success for climate researchers, and he outlined Nvidia's ongoing initiatives to partner with key opinion leaders and researchers through its Earth-2 projects.
Huang noted the first “miracle” would be to mimic the environment quickly and at a high enough resolution, most probably several square kilometers, according to the post on the company’s website.
Second, the use of AI will be used to simulate the workings of the climate system to allow researchers to study future outcomes with speed and great resolution, according to the website, and advances in generative AI offer hope for the development of ways to produce km-scale predictions of the global climate.
The third miracle, Huang opined, according to the website, would feature the use of Nvidia Omniverse to visualize the data in a dynamic way, and this “puts it in the hands of policymakers, businesses, companies, and researchers.”
To provide accessible km-scale climate information for the first time to manage the global climate in a sustainable way, according to the website, the Earth Virtualization Engines initiative (EVE) brings together digital infrastructure focused on climate science. As part of this, AI, ICON, HPC, IFS, MPAS, NEMO, WRF-G, and a range of additional applications benefit from today’s high-speed computing and a boost to this computing power is in the near future.