From climate modeling to protecting endangered species, developers, researchers, and companies are keeping AI in the environment with the help of NVIDIA technology.
They are using NVIDIA GPUs and software to track the endangered African black rhinoceros, predict the availability of solar energy in the UK, build detailed climate models, and analyze the environment from satellite imagery. Watching for disasters.
This Earth Day, discover five key ways AI and accelerated computing are improving sustainability, climate science, and energy efficiency.
1. Application of AI to biodiversity conservation and sustainable agriculture
To protect endangered species, camera-enabled edge AI devices embedded in environments and drones help scientists observe animals in the wild, monitor populations, and detect threats from predators and poachers. help you to
Conservation AI, a UK-based nonprofit, has deployed over 70 cameras worldwide with NVIDIA Jetson modules for edge AI. The Conservation AI platform, combined with the NVIDIA Triton Inference Server, can identify species of interest from video in just four seconds, helping conservationists detect poachers and intervene quickly. Another research team developed an NVIDIA Jetson-based solution and used drone-based AI to monitor the endangered black rhinoceros in Namibia.
Also, artist Sofia Crespo is raising awareness of endangered plant and animal species through an exhibit of generative AI art in Times Square, using generative adversarial networks trained on NVIDIA GPUs to create relatively Created high resolution visuals representing unknown species.
In agriculture, Bay Area startup Verdant and smart tractor company Monarch Tractor are developing AI to support sustainable farming practices, such as precision spraying to reduce herbicide use.
2. Promotion of renewable energy research
NVIDIA AI and high-performance computing are advancing nearly every area of renewable energy research.
Open Climate Fix, a non-profit product lab and member of the NVIDIA Inception Program for Startups, is developing an AI model to help predict cloud cover on solar panels. This helps grid operators determine how much solar energy they can generate on a given day to serve their customers. ‘ Power required. His Utilidata and Anuranet startups are using NVIDIA Jetson to develop AI-enabled electricity meters for a more energy efficient and resilient grid.
Siemens Gamesa Renewable Energy collaborates with NVIDIA to create a physics-based digital twin of a wind farm using NVIDIA Omniverse and NVIDIA Modulus. British company Zenotech used cloud-based GPUs to accurately simulate the estimated energy output of 140 turbines in a wind farm. And a consortium-led project, Gigastack, is using Omniverse to build a proof-of-concept wind farm that turns water into hydrogen fuel.
Researchers at Lawrence Livermore National Laboratory Achieve Fusion Energy Breakthrough Using HPC Simulations Running on Sierra, World’s 6th Fastest HPC System with 17,280 NVIDIA GPUs . And the UK Nuclear Agency is testing the NVIDIA Omniverse simulation platform to design a fusion energy power plant.
3. Accelerate visualization of climate models and weather
Accurate modeling of the atmosphere is important for predicting climate change in the coming decades.
To more accurately predict extreme weather events, NVIDIA created FourCastNet. This is a physical ML model that can predict the exact path of catastrophic atmospheric rivers a week in advance.
NVIDIA and Lockheed Martin are using Omniverse to build an AI-powered digital twin for the National Oceanic and Atmospheric Administration. This can significantly reduce the time required to generate complex weather visualizations.
An initiative by researchers at Northwestern University and Argonne National Laboratory uses NVIDIA Jetson-powered devices to better understand the impact of wildfires, urban heat islands, and climate on crops in hyperlocal approach is adopted.
4. Managing Environmental Disasters with Satellite Data
Satellite data provides a powerful perspective for monitoring and managing climate hazards when it is difficult to grasp the situation from the ground.
NVIDIA is working with the United Nations Satellite Center to apply AI to the organization’s satellite imaging technology infrastructure. This is an initiative that provides humanitarian teams with near real-time insights into floods, wildfires, and other climate-related disasters.
NVIDIA Inception member Masterful AI has developed a machine learning tool that can detect climate risk from satellite and drone feeds. This model is being used to identify rusting transformers that may cause wildfires and improve post-hurricane damage assessments.
San Francisco-based Inception startup Orbital Sidekick operates satellites that gather hyperspectral intelligence — information from across the electromagnetic spectrum. The company’s NVIDIA Jetson-powered AI solution can detect hydrocarbon and gas leaks from this data, reducing the risk of leaks becoming a serious crisis.
5. Promoting Energy Efficient Computing
Adopting NVIDIA technology is already a green choice. If all the CPU-only servers in the world running AI and HPC were switched to GPU-accelerated systems, the world could save about 20 trillion watt hours of energy annually. Power requirements for approximately 2 million US homes.
Semiconductor industry leaders integrate NVIDIA cuLitho software libraries to speed time-to-market and make computer lithography, the next-generation chip design and manufacturing process, more energy efficient. And his NVIDIA Grace CPU Superchip, which recorded a 2x performance improvement over comparable x86 processors in testing, can help cut data center power bills by up to half.
In the latest MLPerf inference benchmark for AI performance, NVIDIA Jetson AGX Orin system-on-module achieved up to 63% improvement in energy efficiency, delivering AI inference at low power levels, including battery-powered systems.
NVIDIA announced the liquid-cooled NVIDIA A100 Tensor Core GPU last year. This has been evaluated by Equinix for use in their own data centers. The companies found that data centers using liquid cooling can run the same workloads as air-cooled facilities while using approximately 30% less energy.
BONUS: AI Podcast Robot-Assisted Recycling
Startup EverestLabs has developed RecycleOS, an AI software and robotics solution that helps recycling facilities around the world recover an average of 25-40% more waste. Here’s what the company’s founder and CEO had to say about his company’s technology on his NVIDIA AI podcast:
Learn more about green computing and NVIDIA accelerated applications in climate and energy.
