Chipmaker Nvidia is building a bridge from to its “digital twin” metaverse technology called omniverse to Earth-2, a project originally built to predict the weather with greater accuracy.
The National Oceanic and Atmospheric Administration (NOAA) is developing a new global weather modeling system with a little help from Lockheed Martin and chip maker Nvidia.
The Earth-2 project aims to predict global weather patterns with far greater accuracy than possible at present and to do so in a far more intuitive and visually-oriented manner.
Lockheed Martin will contribute its OpenRosetta3D software to the project. OpenRosetta3D is the same software NASA uses for large-scale image analysis and geospatial modeling. Nvidia will build a bridge from this software to its “digital twin” omniverse (metaverse).
Nvidia’s omniverse will render the data with Universal Scene Description (USD) – which Nvidia calls the HTML of 3D.
Nvidia expects the Earth-2 supercomputer to greatly increase the fidelity of climate modeling. At present, the very best systems have 10 to 100-kilometer resolutions.
To simulate clouds that reflect sunlight back to space, those resolutions need to be much sharper, right down to the individual meter. That level of advancement would ordinarily take decades to achieve, but Nvidia believes it has found the express lane.
“For the first time, we have the technology to do ultra-high-resolution climate modeling, to jump to lightspeed and predict changes in regional extreme weather decades out,” says Nvidia.
According to the GPU manufacturer, three technologies will power Earth-2. These are:
“GPU-accelerated computing; deep learning and breakthroughs in physics-informed neural networks; and AI supercomputers, along with vast quantities of observed and model data to learn from.”
Data should not be a major hurdle as the NOAA already has vast data sets to employ in the system.
Climate science and more
Nvidia touts Earth-2 as a major potential breakthrough for climate science. As Nvidia sees it, however, the applications of AI and its omniverse technology are manifold.
Ian Buck, the vice president of Accelerated Computing at NVIDIA predicts that the technology will swiftly work its way into other major industries too.
“Accelerated computing with AI at data center scale has the potential to deliver millionfold increases in performance to tackle challenges, such as mitigating climate change, discovering drugs and finding new sources of renewable energy,” said Buck.
The marketing for this new technology stack is certainly on point.
Nvidia has set high expectations for its climate model and has set a delivery schedule for the final quarter of 2023. In the absence of a future-predicting supercomputer, interested parties will simply need to wait and see whether Earth-2 can live up to all of its lofty promises.