Emerald AI and Lumen Partner to Develop Energy-Aware Networking for AI Data Centres

Emerald AI and Lumen Technologies have announced a collaboration to develop energy-aware networking technology that could automatically shift artificial intelligence (AI) computing workloads between data centres based on electricity availability, cost and network conditions. Announced on 29 October 2026, the initiative aims to help AI infrastructure operators use existing electricity grid capacity more effectively while maintaining computing performance. By adjusting where AI workloads run, the companies hope to reduce electricity demand in locations facing grid constraints without waiting for additional power generation or transmission infrastructure.

The collaboration will integrate Emerald AI's Emerald Conductor platform with Lumen's programmable fibre network and edge computing infrastructure. The proposed system will use information on electricity availability, network latency, security and operating costs to determine where AI inference requests should be processed. During periods of high electricity demand, computing tasks could be redirected towards facilities with greater available power capacity. This approach could help data centre operators manage electricity costs, improve operational flexibility and reduce pressure on local grids while continuing to deliver AI services.

 

Read more: Emvolon Raises $22 Million to Scale Modular Waste-to-Fuel and Biomethanol Technology

 

Dr Varun Sivaram, Founder and CEO of Emerald AI, said, "Power is the critical bottleneck to AI growth". He explained that integrating Emerald Conductor with Lumen's network infrastructure could enable AI inference workloads to move between locations while maintaining low latency and high performance. Jim Fowler, Chief Technology and Product Officer at Lumen, said, "We’re exploring how to turn that equation around, making the network an active part of how and where computing happens." The collaboration extends Emerald AI's existing power management approach beyond individual data centres into the wider communications network connecting AI infrastructure.

The companies are exploring a pilot using Lumen's edge locations and network infrastructure to test automated workload routing under different electricity and network conditions. Kate Johnson, CEO of Lumen, highlighted the company's broader efforts to make networking infrastructure more responsive to the changing requirements of AI computing. Although Emerald AI estimates that flexible data centres could potentially unlock up to 100 gigawatts of existing US grid capacity, this represents a company-stated opportunity rather than capacity demonstrated through the collaboration. No pilot results, verified energy savings or emissions reductions have yet been disclosed.

 

Source: Emerald AI