At the “Building a New Paradigm for Integrated Development of AI Computing Power and Green Energy” Computing Collaboration Symposium held on the first day of the 2026 World Artificial Intelligence Conference, Shangtang University released a new “Computing Collaborative Agent” and pioneered the proposal to redefine the AIDC efficiency scale from PUE to TPW to promote system-level cost reduction and efficiency of intelligent computing centers. At present, the Computing and Computing Collaboration Agent has been fully implemented at AIDC in Lingang, Shangtang, to achieve an 80% increase in unit electricity cost token output, 10% lower than IDC in the same region, and a computing power load prediction accuracy of 96%. It is estimated that carbon emissions can be reduced by 24,000 tons/10,000 P for the whole year, driving AIDC from traditional energy management to system-level computational efficiency optimization. Relying on the standardized delivery process of the FDE expert system of Shangtang University, the deployment and launch can be completed as soon as one week.

At the “Building a New Paradigm for Integrated Development of AI Computing Power and Green Energy” Computing Collaboration Symposium held on the first day of the 2026 World Artificial Intelligence Conference, Shangtang University released a new “Computing Collaborative Agent” and pioneered the proposal to redefine the AIDC efficiency scale from PUE to TPW to promote system-level cost reduction and efficiency of intelligent computing centers. At present, the Computing and Computing Collaboration Agent has been fully implemented at AIDC in Lingang, Shangtang, to achieve an 80% increase in unit electricity cost token output, 10% lower than IDC in the same region, and a computing power load prediction accuracy of 96%. It is estimated that carbon emissions can be reduced by 24,000 tons/10,000 P for the whole year, driving AIDC from traditional energy management to system-level computational efficiency optimization. Relying on the standardized delivery process of the FDE expert system of Shangtang University, the deployment and launch can be completed as soon as one week.