Nvidia wants to create a digital twin of planet Earth in the “omniverse” and use this model to glimpse the future of climate change. With this digital twin, the company hopes to see future human challenges in terms of global warming and the accompanying natural disasters, and provide early warnings to regions, nations and cities that may be most affected.
The company has announced plans to build a new AI supercomputer, dubbed Earth-2 (Earth-2, or E-2) for short, and a system inside it to predict climate change. The company says the idea is no longer in the realm of science fiction because of advances in three key technologies: GPU-accelerated computing; deep learning; Model data treasures to be observed and learned with physics-informed neural networks and artificial intelligence supercomputers.
Company CEO Jensen Huang confirmed that the project was fully funded by Nvidia and said that no other company, nation-state or international organization has yet participated: “We now have several supercomputers, the largest in the world. Selene is number six in the world. Cambridge-1 (health research “The AI supercomputer used for healthcare has started operating as a supercomputer dedicated to healthcare – about the size of Selene. So we have three or four supercomputers about the size of the world’s sixth largest supercomputer. And it’s fully funded by us.”
Huang added that their aim is to help in the fight against climate change: “We did (this project) so that we could develop our software in the modern artificial intelligence world. The software was developed by great computer scientists, but accelerated and strengthened by the software development of machines. Earth-2. We will create to focus on the ability of our researchers to contribute to the fight against climate change.”
Huang continued, saying that they plan to introduce Earth-2’s architecture soon, as it will allow the company to create “the most energy-efficient supercomputer ever developed.”
The problem with predicting climate change, as opposed to forecasting the weather, is that the time interval is too long and there are too many variables to consider. Nvidia argues that it needs climate models that can predict the climate in different parts of the world over decades to mitigate the effects of climate change and develop the best strategies to adapt to changes.
These models appear as “perennial simulations” that model the physics, chemistry and biology of the atmosphere, waters, ice, soil and human activities. But all this remains superficial, because in addition to the need to simulate clouds reflecting sunlight back into space, there is also a need to simulate the global water cycle (water movement from the ocean, sea ice, land surface and groundwater, atmosphere and clouds).
Nvidia concludes that it has the technology to perform ultra-high-resolution climate modeling for the world’s first time. . .
33861
