NVIDIA is building Earth's digital 'twin' to fight climate change

At the recent GTC event, Nvidia unveiled the Earth-2 project, the world's most powerful AI supercomputer, a 3D simulator of the Earth with flexible and detailed simulation capabilities to help people evaluate the real world. more accurately about climate change as well as predicting future disasters.

The earth is getting hotter and hotter. The past 7 years can be considered as the 7 hottest years in human history. Since the period 1850 - 1900, greenhouse gas emissions from human activities have caused the average annual temperature of the blue planet to increase by about 1.1 degrees Celsius. What we are experiencing. really was at an alarming rate, with a series of extreme weather conditions such as prolonged drought, heavy rain, storm, hurricane and terrible floods. Natural disasters that were already harsh have become more cruel and intense than ever.

However, overcoming climate change is not simple and cannot be done overnight. Just a few years, or even a few decades, is not enough for us to feel positive changes. However, in order for the campaigns to save the Earth and save the environment to take place in the right way, we need to see the future, to feel what will happen with the highest accuracy.

NVIDIA wants to use Earth's "digital twin" to predict the future.

It sounds fictional, but that's what NVIDIA is after. At the GTC event in mid-November, they unveiled the Earth-2 (E-2) project, the world's most powerful AI supercomputer, the 'twin brother' of our Earth. on the Omniverse platform, a 3D simulator of the blue planet with the ability to make more accurate predictions about climate change.

According to CEO Jensen Huang, E-2 will use a combination of three high-end technologies: high-speed GPU, deep learning, and AI supercomputer, along with a huge database of Earth. All will aid in the creation of the most accurate and super high quality resolution climate prototypes.

In fact, there are already many models like the E-2 in existence, with the ability to determine factors such as air pressure, wind strength and temperature to create the appropriate equations, giving a view. objectively about climate patterns in certain regions. As that area will be represented as a 3D mesh. The smaller the area, the more accurate the calculations will be, and the more realistic the simulator will be before it becomes unusable.

In other words, weather models need to solve more equations to achieve higher resolution. However, the more equations you take in, the slower the model becomes, less efficient, and gradually no longer useful. That's exactly the problem most climate models have today: Lacking both detail and accuracy.

And the solution that NVIDIA offers is a bigger, better, faster supercomputer. On the company blog, Mr. Huang wrote: 'We need higher resolution to simulate changes in the global water cycle. We need meter-scale resolution to simulate the process of sunlight being reflected by clouds, bounced back into space. Scientists calculate that such resolutions require millions to billions of times more computing power than we have at the moment'.

Earth-2 will solve problems that exist in other similar emulators.

Going back to E-2, Earth's digital twin was created with the purpose of promoting actions that both alleviate climate change while also minimize its negative impacts on nature and people. Extreme weather events such as storms, wildfires, heat waves or flash floods are becoming more unpredictable, with more devastating effects.

That situation would be somewhat improved if we could more accurately predict such disasters in the future. Mr. Huang hopes NVIDIA's model can predict decades of extreme weather changes in many regions of the world. At that time, people will have time to come up with timely response solutions, reasonable evacuation policies, or appropriate construction projects to adapt well in such climate conditions.

E-2 is also used to find a reasonable solution, simulating many different plans to find out which is the most effective, with the lowest cost. This is also considered one of NVIDIA's biggest projects in many years. Mr. Huang said: 'All the technologies that we have created so far are necessary to be able to make E-2 a reality. I'm really looking forward to new features, more important tasks that this model can take on in the future."

It's not just about climate change

The combination of different technologies has helped NVIDIA create the most advanced and efficient Earth simulation, and has also helped them solve many problems related to supercomputer speed, especially especially in research projects with huge databases.

Similar to E-2, NVIDIA has focused on developing 3 core technologies: High-performance computing, AI, and the scale of data centers. This not only helped them simulate the Earth, but also created many digital twins of cities and factories around the world. This large-scale simulation technology is still very new and also full of potential.

In addition, the team at NVIDIA, Cal Tech and start-up Entos have successfully combined machine learning with physics to create the OrbNet program. As a result, Entos was able to speed up their new drug discovery simulator 1000x, completing a workload over 3 months in as little as 3 hours.