Electricity changes the AI competition landscape! Goldman Sachs tested: By 2030, America's "backup power" will fall below the critical line, while China still has up to 400GW
Goldman Sachs expects that by 2030, the "effective backup power capacity" in the United States is projected to fall below the industry critical threshold of 15%, especially in data center-intensive areas, which may become a bottleneck for AI development. In contrast, thanks to large-scale power infrastructure construction, China is expected to have approximately 400 gigawatts (GW) of backup power by 2030, more than three times the anticipated global demand from data centers
Electricity - Its Role in the AI Competition is Becoming Increasingly Important.
According to news from the Chasing Wind Trading Desk, a Goldman Sachs research report dated November 13 indicates that the significant disparity in electricity supply is becoming a key variable that may slow down the expansion of the United States in the AI infrastructure sector while providing China with an unexpected competitive advantage.
As the demand for computing power from AI explodes, the United States is facing an imminent electricity supply crisis. Goldman Sachs estimates that by 2030, under a scenario of rapid growth in data center demand, the effective reserve electricity capacity across the United States may fall below the industry-recognized "critical line" of 15%, which means that the reliability of the power grid will face severe challenges.
In contrast, the situation in China is entirely different. The report predicts that thanks to large-scale investments in renewable energy, coal power, nuclear power, and other fields in recent years, China will have approximately 400 gigawatts (GW) of substantial reserve electricity capacity by 2030. This figure not only far exceeds its own demand but is also more than three times the expected total electricity demand of global data centers at that time (approximately 120GW).
Analysts Hongcen Wei, Daan Struyven, and Samantha Dart emphasize in the report that sufficient and reliable electricity supply is a key shaping factor in the AI competition.
U.S. Electricity Alarm: Reserve Capacity Approaching "Red Line"
Goldman Sachs' report sounds the alarm for the electricity outlook in the United States. Currently, the U.S. holds 44% of the world's data center capacity, leading the way, but its power grid is under increasing pressure. Data from the report shows that data centers account for about 6% of the total electricity demand in the United States, and this proportion is expected to rise to 11% by 2030.
Even more concerning is that the "effective reserve electricity capacity" in the U.S.—a key indicator of the reliability of the power grid during peak electricity usage—has significantly dropped from 26% to 19% over the past five years, gradually approaching the critical threshold of 15%.
At the regional level, the problem is even more pronounced: among the 13 regional power grids in the U.S., 8 are at or below this critical level. This tense situation has already caused substantial impacts in some areas, such as the PJM power market in Virginia, known as the "data center capital of the world," which experienced significant spikes in real-time electricity prices and generation capacity prices last summer.
Goldman Sachs analysts warn: "Limited effective reserve electricity capacity may become a bottleneck for the further development of data centers in the United States."
China's Abundant Electricity: 400GW of Redundancy Provides Development Space
In stark contrast to the tightness of the U.S. power grid, China's electricity supply appears very ample. According to Bloomberg, after experiencing power shortages in 2021 and 2022, China has embarked on a large-scale electricity construction initiative covering various energy forms, including renewable energy, coal power plants, and nuclear reactors.
Goldman Sachs' report quantifies this construction achievement. The report predicts that by 2030, China's effective reserve electricity capacity will reach 25% of peak demand, corresponding to an absolute value of approximately 400GW. This enormous electricity redundancy provides a solid foundation for China to undertake the future AI-driven expansion of data centers and the development of other energy-intensive industries"We expect China's backup capacity to remain sufficient to accommodate the growing electricity demand from data centers and other industries," Goldman Sachs analysts wrote in the report. Currently, China holds about a quarter of the world's data center capacity, ranking second globally. Abundant power reserves may enable it to quickly catch up with the leaders.
Supply and Demand Imbalance: Analyzing the Divergence in Power Dynamics between China and the U.S.
The significant divergence in power dynamics between China and the U.S. stems from different paths taken on both the supply and demand sides.
The Goldman Sachs report analyzes that the tightening of the U.S. electricity market is primarily driven by three factors: first, robust growth in electricity demand driven by data centers; second, insufficient construction of new generation capacity, with new renewable energy and natural gas installations unable to offset planned coal retirements; and finally, the scale of new technologies such as energy storage remains limited, unable to fill the gap.
In contrast, in China, the government has promoted comprehensive investment in all power sources for strategic considerations of energy security. The report points out that this round of construction is not only large in scale but also encompasses various technologies including renewable energy, coal, natural gas, and nuclear power, aiming to systematically enhance the backup capacity and reliability of the power grid.
Electricity: A New Variable in the AI Race
This Goldman Sachs report reveals a long-term yet crucial new variable in the AI race. In the past, the market's focus was primarily on chip bans, talent mobility, and algorithm innovation, while the importance of electricity as a physical foundation is becoming increasingly prominent.
The report concludes that bottlenecks in electricity infrastructure may temporarily slow down AI development in the U.S. at least until 2030. For companies and investors relying on large-scale data centers, there may be higher electricity costs, longer project approval times, and even risks of power shortages in certain regions of the U.S. in the future.
Conversely, China's abundant and controllable electricity supply may become a new magnet for attracting data center investments. This trend could not only change the geographical distribution of global data centers but also have far-reaching impacts on the upstream and downstream industrial chains, including energy equipment and energy-intensive industries (such as aluminum).
The above content is from [追风交易台](https://mp.weixin.qq.com/s/uua05g5qk-N2J7h91pyqxQ).For more detailed interpretations, including real-time analysis and frontline research, please join the【 [追风交易台▪年度会员](https://wallstreetcn.com/shop/item/1000309)】[](https://wallstreetcn.com/shop/item/1000309)