UHV Grid vs. America's Aging Infrastructure
For years, the AI race has been framed as a competition between chips. Nvidia vs. AMD. U.S. vs. China. Silicon vs. silicon. But there is another gap that may matter more. It has nothing to do with fabrication nodes.
The U.S. electric grid was built in the 1960s and 1970s. The average large power transformer in service today is 38 to 40 years old. Over 70% of U.S. transmission lines are more than 25 years old — past the halfway point of their operational lifespan. The average age of the grid itself is pushing 60 years in many areas.
China, meanwhile, built the world's largest hybrid power grid. It did so in just two decades. It has constructed 46 ultrahigh-voltage "power highways" that span the country. Its west-to-east transmission capacity has reached 340 million kilowatts. And it has done all of this while the U.S. was still debating whether to upgrade its own infrastructure.
This is the infrastructure gap that nobody is talking about. And it may be one of the most consequential gaps in the AI race.
What China Built
China's grid transformation is unlike anything the world has ever seen. The country has built 46 UHV projects as of the end of 2025. They include 22 AC and 24 DC lines. Total length exceeds 62,000 kilometers — enough to circle the Earth one and a half times. By 2026, that figure is expected to surpass 70,000 kilometers.
China is the only country in the world to achieve large-scale commercial operation of UHV transmission. Its ±1,100 kV DC line holds the global record for the highest voltage, largest capacity, and longest distance. The Changji-Guquan project delivers 12 million kilowatts over 3,300 kilometers. Transmission losses on these lines are just 2.8% to 4%. Europe and North America average approximately 7%, according to industry data.
The system moves electricity from the resource-rich west to the energy-hungry east. Twenty-four DC "power arteries" are now in operation. The west-to-east transmission capacity reached 340 million kilowatts in 2025. It is expected to exceed 420 million kilowatts by 2030. China's renewable energy utilization rate has stayed above 97% for six consecutive years.
All of this is backed by record investment. The "15th Five-Year Plan" (2026-2030) calls for at least 5 trillion yuan (approximately $700 billion) in grid investment. State Grid alone is expected to invest 4 trillion yuan (around $570 billion) during the period. The plan includes 15 new UHV DC transmission lines. Inner Mongolia alone will build 12 UHV projects.
China has built a national power grid that moves electricity with minimal loss. It integrates massive renewable energy. It has maintained zero large-scale outages for years. It has turned a fundamental constraint into a solved problem. The constraint: resources in the west, demand in the east.
What the U.S. Hasn't Built
The U.S. story is almost the reverse.
The American grid was largely constructed in the 1960s and 1970s. It was designed for a different era — one with stable demand, predictable growth, and no AI. Today, 70% of transmission lines and power transformers are more than 25 years old. Many are operating beyond their intended lifespan.
The average large power transformer in the U.S. is 38 to 40 years old. Failure probability rises sharply once a unit passes 30 to 40 years. Roughly 55% of distribution transformers are over 33 years old. The system is aging faster than it is being replaced.
The investment gap is enormous. The U.S. grid needs an estimated $2.5 trillion in modernization, but available capital falls far short. Investor-owned utilities are projected to spend more than $170 billion on transmission and distribution in 2026. Even that is a fraction of what is needed. At current replacement rates of less than 1% per year, it would take decades to modernize the system.
The result is a grid that is not just old, but increasingly unreliable. JPMorgan has labeled the aging U.S. power grid a "national security risk," warning that decades-old infrastructure is vulnerable to extreme weather, cyberattacks, and geopolitical threats. The Department of Energy projects AI data centers could require 100 extra gigawatts of peak capacity. The target year is 2030. Plans for U.S. data center additions fell by half in the fourth quarter of 2025. One reason: the grid is nearing its limit.
The U.S. is not building the infrastructure it needs to power the AI future. It is patching a system designed for the 20th century while trying to compete in the 21st.
The Gap That Matters
The contrast is stark.
China has built a grid that can move electricity across the country with less than 4% loss. The U.S. has a grid that loses roughly 7% over much shorter distances.
China has 46 UHV lines spanning 62,000 kilometers. The U.S. is just beginning to build its first major 765 kV "superhighway" lines.
China plans to add 15 more UHV lines by 2030. The U.S. has interconnection queues that stretch four to seven years.
The Global Power Development Index 2026 gave China a score of 98.1 in technological innovation. That was the highest in the world. The U.S. scored 94.5. The index was released at the UN in July.
This gap matters for AI. The chips that power AI are useless without electricity. The data centers that train models are useless without a grid that can deliver power reliably. The U.S. has the best AI companies in the world. It does not have the infrastructure to power them.
China may still trail the U.S. in chip design. But it has built the grid that the U.S. still cannot. And in the AI race, that gap is worth paying attention to.
Sources: U.S. Department of Energy; JPMorgan grid resilience report (March 2026); Global Power Development Index 2026 (GEIDCO/UN, July 2026); State Grid Corporation of China; National Energy Administration of China; China Electricity Council; China Daily; Borneo Bulletin; State Council Report on UHV construction (2026).
Disclaimer
The information provided in this article is for general informational and educational purposes only. It does not constitute legal, financial, or professional advice. The author and publisher are not responsible for any actions taken based on the content of this article. Readers should consult qualified professionals for advice specific to their situation. All trademarks and references to third-party products, services, or organizations are the property of their respective owners. The performance data and benchmarks discussed are based on specific research studies and may not generalize to all use cases or environments. As of the publication date, the AI landscape continues to evolve rapidly, and readers should verify current information independently.
Limitations
This analysis is based on reporting and public data available as of the article date; figures may be revised as sources update.
Forecasts from third-party analysts can change with market conditions.
Cost and pricing examples are point-in-time estimates; actual rates vary.
Country and company comparisons rely on public reporting, not operational data.
This sector moves fast; timelines and deal terms may be updated later.
Company deals and regulatory rulings may evolve; verify current status.
AI infrastructure is changing quickly; claims can become outdated soon.
Sources
- U.S. Department of Energy
- JPMorgan grid resilience report (March 2026)
- Global Power Development Index 2026 (GEIDCO/UN, July 2026)
- State Grid Corporation of China
- National Energy Administration of China
- China Electricity Council
- China Daily
- Borneo Bulletin
- State Council Report on UHV construction (2026).
The information provided in this article is for general informational and educational purposes only. It does not constitute legal, financial, or professional advice. The author and publisher are not responsible for any actions taken based on the content of this article. Readers should consult qualified professionals for advice specific to their situation. All trademarks and references to third-party products, services, or organizations are the property of their respective owners. The performance data and benchmarks discussed are based on specific research studies and may not generalize to all use cases or environments. As of the publication date, the AI landscape continues to evolve rapidly, and readers should verify current information independently.
Limitations: This analysis is based on reporting and public data available as of the article date; figures may be revised as sources update.; Forecasts from third-party analysts can change with market conditions.; Cost and pricing examples are point-in-time estimates; actual rates vary.; Country and company comparisons rely on public reporting, not operational data.; This sector moves fast; timelines and deal terms may be updated later.; Company deals and regulatory rulings may evolve; verify current status.; AI infrastructure is changing quickly; claims can become outdated soon.