McKinsey forecasts that by 2030, the global demand for data center capacity will triple, with AI workloads driving 70% of this new demand. Among them, generative AI is expected to contribute roughly 40%. This rapid expansion will propel global data center investment to a staggering $6.7 trillion, yet it will also bring about significant challenges related to electricity, water resources, and land use. In the United States, the electricity consumption of data centers is projected to surge from 17 gigawatts in 2022 to 35 gigawatts by 2030, accounting for 7.5% of the nation's total power output. When it comes to water usage, a single hyperscale data center can gulp down up to 500,000 gallons of water daily. For instance, Meta's project in Georgia has already resulted in local water scarcity. The expansion of data centers in Europe is also under resource strain. By 2028, data center electricity consumption in Ireland could potentially make up 27% of the national power supply. Meanwhile, the Netherlands has had to revise its data center location plans due to nitrogen emission limits. From a technological standpoint, the power density of server racks in AI data centers has doubled over the past two years and is anticipated to quadruple by 2027, further intensifying the demands for energy and cooling.