IT Brief Asia - Technology news for CIOs & IT decision-makers
Story image

NVIDIA launches blueprint to revolutionise weather forecasts

Yesterday

NVIDIA has introduced the NVIDIA Omniverse Blueprint for Earth-2 weather analytics, aiming to advance the development of more precise weather forecasting solutions.

Over the past decade, climate-related weather events have resulted in an economic impact amounting to USD $2 trillion globally. In response, the Omniverse Blueprint is designed to enhance organisations' abilities in risk management and preparedness for disasters using current technological advancements.

The Omniverse Blueprint offers a comprehensive set of reference workflows, including NVIDIA GPU acceleration libraries, a physics-AI framework, development tools, and microservices. These tools facilitate enterprises in transitioning from the prototyping phase to full-scale production with weather forecast models.

Key features of the blueprint are easy-to-deploy NVIDIA NIM microservices for NVIDIA Earth-2, such as CorrDiff for downscaling and FourCastNet for global atmospheric dynamics predictions. These are already employed by various weather technology firms, researchers, and governmental bodies to gather insights and reduce risks associated with severe weather.

"We're seeing more extreme weather events and natural disasters than ever, threatening lives and property," stated Jensen Huang, Founder and CEO of NVIDIA. "The NVIDIA Omniverse Blueprint for Earth-2 will help industries around the world prepare for — and mitigate — climate change and weather-related disasters."

Prominent climate tech companies like AI firm G42, JBA Risk Management, and Spire are adopting the blueprint for the creation of tailored AI-augmented solutions. Within the USD $20 billion climate tech sector, the NVIDIA Earth-2 platform enhances developers' capacity to issue warnings and update forecasts swiftly, often in seconds compared to the minutes or hours required by traditional CPU-based models.

G42 is incorporating various components of the Omniverse Blueprint into its advanced forecasting models for Earth-2 to provide the UAE's National Center of Meteorology with cutting-edge AI technologies for sophisticated weather forecasting and emergency management.

"G42 is advancing AI-powered forecasting to help governments and enterprises strengthen resilience against extreme weather in a rapidly changing world," remarked Andrew Jackson, CEO of Inception, a G42 company. "Using high-resolution weather and climate modeling, we are transforming how organisations anticipate and respond to severe weather conditions with precision and speed. Building on NVIDIA's CorrDiff model, we have developed a custom AI-driven system that downscales coarse weather data into hyper-local forecasts, enabling faster predictions at an unprecedented scale. Combined with the Earth-2 Blueprint, this technology equips decision-makers with actionable intelligence to protect communities, safeguard infrastructure and plan for a more resilient future."

Spire Global utilised AI components from the blueprint as a basis for developing new AI products that combine its proprietary satellite data to offer medium-range and sub-seasonal forecasts up to 45 days. Spire's models, driven by NVIDIA GPUs and the Omniverse Blueprint, operate significantly faster, 1,000 times faster than conventional physics-based models, allowing for large-scale ensemble forecasts capturing extensive weather eventualities.

Organisations like the Central Weather Administration of Taiwan and The Weather Company, alongside companies like 3D mapping specialist Ecopia, spatial analytics firm ESRI, green energy provider GCL Power, JBA Risk Management, aerospace enterprise OroraTech, and the resilience platform Tomorrow.io, are either employing or exploring the utilisation of Earth-2.

The Earth-2 platform encompasses tools, microservices, and an assortment of advanced AI weather models for the visualisation and simulation of global patterns. CorrDiff, a part of the Omniverse Blueprint, can deliver precise weather forecasts much faster and more energy-efficiently than CPUs.

The Omniverse Blueprint for Earth-2 facilitates independent software vendors in developing and deploying AI-enhanced solutions using observational data to improve accuracy and speed. Esri collaborates with NVIDIA to link its ArcGIS platform to Earth-2, while OroraTech considers connecting its data platform to the blueprint.

Tomorrow.io has contributed its near-real-time proprietary satellite data, aiding NVIDIA in creating a digital twin of Earth for training, inferring, and reinforcing next-generation AI models.

A distinct feature of the blueprint is NVIDIA Omniverse, a platform for the development of 3D workflows and applications based on OpenUSD. This blueprint demonstrates how developers can use NVIDIA RTX-powered visualisation pipelines for rendering geospatial and weather data through Omniverse SDKs and microservices.

The Omniverse Blueprint for Earth-2 utilises the NVIDIA DGX Cloud platform for full-stack acceleration in AI-enhanced weather forecasting. Operating on NVIDIA's supercomputers, it offers a pathway to swiftly simulate and visualise global climate patterns on a large scale.

Follow us on:
Follow us on LinkedIn Follow us on X
Share on:
Share on LinkedIn Share on X