Nvidia and EPRI Launch Open Power AI Consortium to Transform the Power Industry
NVIDIA announced the formation of the Open Power AI Consortium, in partnership with the Electric Power Research Institute (EPRI), to revolutionize the world's power industry using artificial intelligence (AI). The consortium unites energy companies, technology companies, and researchers to create AI solutions that solve critical issues in the power sector, including enhancing grid reliability, optimizing asset performance, and improving energy management.
A New Era for Power and Energy Management
The Open Power AI Consortium is led by the EPRI, an independent, nonprofit R&D organization devoted to energy. The effort works to speed AI adoption in the electric power sector by developing open AI models using high-quality, power sector-specific data.
EPRI’s president and CEO, Arshad Mansoor, said, “Over the next decade, AI has the great potential to revolutionize the power sector by delivering the capability to enhance grid reliability, optimize asset performance, and enable more efficient energy management.”
Also read: NVIDIA, Google DeepMind, and Disney Launch Newton, an Open-Source Physics Engine
EPRI, NVIDIA, and Articul8 are creating multimodal language models specifically for the energy industry, trained on EPRI's private data, to optimize utility operations, increase energy efficiency, and make grids more resilient.
The first publicly available AI model for power and electric systems, developed with hundreds of NVIDIA H100 GPUs, will soon be available for early access as an NVIDIA NIM microservice.
The consortium, backed by over 20 large energy companies and technology companies such as AWS, Oracle, and Microsoft, plans to grow globally and drive AI adoption in the energy industry. As energy consumption increases, AI technologies will enable utilities to enhance infrastructure resilience, balance energy supply and demand, and optimize regulatory and environmental processes.
Faster, Smarter Infrastructure Development with AI
The consortium seeks to shorten the timeline for interconnection studies, traditionally four years or more, using AI to eliminate the timeframe at least fivefold. AI is also expected to speed up licenses, permits, and utility rate case preparation for energy companies so that they can provide for increased demand while maintaining reliability in the grid.
Standardizing AI Benchmarks for the Power Sector
Apart from publishing data models, the consortium aims to develop a uniform framework of benchmarks for testing the performance and integrity of AI technologies in the power industry. This will assist utilities, researchers, and other industry players in testing the efficiency of AI solutions and making their incorporation into the power sector secure and efficient.
