RETRACTED: Power for AI Data Centers: Energy Demand, Grid Impacts, Challenges and Perspectives

The demand for computing power has increased at a rate never seen before due to the quick development of artificial intelligence (AI) technologies and applications. Consequently, AI data centers, referring to computing facilities specifically designed for large-scale artificial intelligence workloads, have become one of the fastest-growing electricity consumers globally. Therefore, it is essential to understand the load characteristics of AI data centers and their impact on the grid. This paper provides a comprehensive review of the evolving energy landscape of AI data centers. Specifically, this paper (i) presents the energy consumption structure in AI data centers and analyzes the key workload features and patterns in four stages, emphasizing how high power density, temporal variability, and cooling requirements shape total energy use, (ii) examines the impacts of AI data centers for power systems, including impacts on grid stability, reliability and power quality, electricity markets and pricing, economic dispatch and reserve scheduling, and infrastructure planning and coordination, (iii) presents key technological, operational and sustainability challenges for AI data centers, including renewable energy integration, waste heat utilization, carbon-neutral operation, and water–energy nexus constraints, (iv) evaluates emerging solutions and opportunities, spanning grid-side measures, data-center-side strategies, and user-side demand-flexibility mechanisms, (v) identifies future research priorities and policy directions to enable the sustainable co-evolution of AI infrastructure and electric power systems. The review aims to support utilities, system operators, and researchers in maintaining reliable, resilient, and sustainable grid operation in the context of the rapid development of AI data centers.