NVIDIA and Energy Titans Unveil 'Flexible AI Factories' to Power the Future

In a major shift for the technology and energy sectors, NVIDIA and Emerald AI announced a strategic partnership during CERAWeek 2026 with six of the largest power companies in the United States, including AES, Constellation, and NextEra Energy. This collaboration aims to deploy a new class of 'Flexible AI Factories' designed to resolve the growing power bottleneck facing the artificial intelligence industry.
As AI workloads demand unprecedented amounts of electricity, these next-generation facilities will utilize the NVIDIA Vera Rubin DSX reference design. Unlike traditional data centers that act as passive loads on the power grid, these AI factories are engineered to function as grid-responsive assets. Using the DSX Flex software library, they can fluctuate their power consumption in real-time or even supply electricity back to the grid using co-located energy generation and storage systems.
NVIDIA CEO Jensen Huang emphasized that energy, compute, and networking must now be designed as a single, unified architecture. The project claims it could unlock up to 100 gigawatts of stranded grid capacity across the U.S. by allowing AI infrastructure to connect to the power system faster and more efficiently. This 'power-flexible' approach ensures that AI scaling does not compromise grid stability for local communities.
Simultaneously, the cloud ecosystem is adapting to this new hardware reality. Amazon Web Services (AWS) recently announced the integration of NVIDIA's Nemotron 3 Super into its Bedrock platform, while Oracle launched a specialized AI Customer Excellence Centre in Sydney to bridge the gap between regional infrastructure and global AI demand.
Key Sources
- NVIDIA Official Press Room: Flexible AI Factories as Grid Assets
- AWS News Blog: Weekly Roundup March 23, 2026
- Investing.com: NVIDIA Partners with Energy Firms on Grid-Flexible AI Factories
Source Relevance Justification
- NVIDIA (Tier 1): Direct primary source for the Vera Rubin architecture and the Emerald AI partnership details.
- AWS (Tier 1): Primary source for cloud-side implementation of NVIDIA's latest AI models on March 23, 2026.
- Investing.com (Tier 2): Provides validated financial context and broader industry implications for the partnership.