Goldman Sachs(US:GS) 智通财经网·2025-09-03 07:29Core Insights - The rapid rise of artificial intelligence (AI) is reshaping global energy demand, with data centers at the center of this transformation [1] - Goldman Sachs predicts that by 2030, AI data centers will drive global electricity demand to increase by 165% compared to 2023 levels [1] - The construction spending on data centers in the U.S. has doubled in three years due to hyperscale companies accelerating their development to meet AI needs [1] Group 1: Electricity Demand and Data Centers - Global electricity consumption is currently about 55 gigawatts (GW), with over half used for cloud computing, and AI currently accounts for 14% [1] - By 2027, total consumption is expected to reach 84 GW, with a significant increase in AI's share [1] - Average power density in data centers is projected to rise from 162 kilowatts per square foot to 176 kilowatts by 2027 [2] Group 2: Investment and Infrastructure Challenges - Meeting the growing electricity demand will require unprecedented investment, with U.S. utilities needing to add $50 billion in generation capacity for data centers [3] - Global grid upgrade costs could reach $720 billion by 2030 [3] - 40% of new electricity capacity is expected to come from renewable sources, with wind and solar being more cost-competitive than natural gas [3] Group 3: Nuclear Power and Sustainability - Nuclear power is regaining favor as tech companies seek reliable, low-carbon baseload options, with over 10 GW of new nuclear capacity contracts signed in the U.S. alone [3] - Political shifts in markets like the U.S. and Switzerland are opening doors for new nuclear reactor construction [3] Group 4: Efficiency and Emission Reduction - Cooling systems, which account for up to 40% of energy consumption in hyperscale operations, will remain a key focus for efficiency improvements [4] - Long-term strategies involving new power sources and infrastructure could lead to significant reductions in data center emissions intensity [4]