📊 Full opportunity report: The gigawatt gap. Why China is structurally positioned for AI power and the US is engineering around its grid. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.

China has established a structural advantage in powering AI data centers by deploying extensive renewable energy and high-voltage transmission infrastructure, allowing it to operate at gigawatt-scale capacity without the grid constraints faced by the US. This positions China to potentially outpace the US in large-scale AI deployment, despite having less advanced individual chips. Learn more about China’s AI infrastructure strategies.

According to Thorsten Meyer, the US dominates AI chip technology, models, and applications, but it faces significant constraints at the physical infrastructure layer needed to deliver power to data centers. New frontier AI data centers in the US now require 100 MW to start and up to 2 GW at full buildout, with the largest projects targeting 12 GW. The US infrastructure relies heavily on off-grid gas turbines, nuclear contracts, and regulatory arbitrage, leading to long interconnection queue times.

In contrast, China’s approach is centered on the Eastern Data Western Compute initiative, routing demand to western renewable hubs via over 40,000 kilometers of ultra-high-voltage (UHV) transmission lines capable of 340 GW capacity. In 2025, China added over 430 GW of wind and solar, pushing renewable capacity above 1.8 TW, and total capacity to 3.89 TW. Chinese AI chips, like Huawei’s Ascend 910C, perform at about 60% of US chips like the NVIDIA H100, but China compensates for lower chip performance through massive power availability, enabled by its centralized planning and extensive renewable infrastructure.

This structural difference means China can deploy less powerful chips across a vast, renewable-powered grid, effectively substituting raw power for chip performance, whereas the US focuses on optimizing chip efficiency and performance per watt.

Implications of the Gigawatt-Scale Power Divide

This structural divergence could reshape global AI leadership. While the US maintains technological superiority at the chip level, China’s ability to scale AI infrastructure through renewable energy and centralized planning offers a different path to deploying large-scale AI systems. If the US cannot overcome grid and permitting bottlenecks, its future AI deployment might be limited by physical infrastructure constraints, potentially capping its global dominance.

Understanding this gigawatt gap is critical for policymakers and industry leaders, as it highlights that AI capacity at scale is increasingly dependent on infrastructure and energy policy, not just chip performance or model innovation.

Historical and Structural Foundations of US and China AI Infrastructure

The US has historically led in AI chip development, with a fragmented infrastructure system relying on off-grid power sources, regulatory arbitrage, and complex interconnection queues. Major projects like Meta’s Hyperion and OpenAI’s Stargate operate at multi-gigawatt scales but face grid bottlenecks that limit expansion.

China’s centralized planning, driven by the NDRC and State Grid, has prioritized renewable energy and extensive transmission networks, enabling the country to build gigawatt-scale data centers with less concern for local permitting or grid constraints. In 2025, China’s renewable capacity expansion far outpaced the US, supporting its infrastructure-driven approach to AI deployment.

“The gigawatt gap does not stem from chip technology but from the structural differences in how the US and China build and operate their power infrastructure.”

— Thorsten Meyer

Unresolved Questions on Infrastructure and Future AI Capacity

It remains unclear whether the US can overcome its grid and permitting constraints through regulatory reform, technological improvements, or new infrastructure investments. Additionally, it is uncertain how quickly China’s renewable expansion and transmission infrastructure can scale further to support even larger AI deployments. The long-term impact of these structural differences on global AI leadership is still developing.

Next Steps in Infrastructure Development and Policy Responses

Expect continued investment in US grid modernization, regulatory reforms, and innovative energy solutions to address bottlenecks. Meanwhile, China’s ongoing renewable expansion and transmission projects will be monitored for their ability to sustain gigawatt-scale AI data centers. The coming 24 months will be critical in determining whether the US can close the gigawatt gap or if China’s infrastructure advantage becomes a lasting strategic lead.

Key Questions

Why does the US struggle with AI infrastructure expansion?

The US faces regulatory, permitting, and grid constraints that make it difficult to rapidly expand high-capacity power connections needed for gigawatt-scale AI data centers.

How does China’s approach differ from the US in powering AI data centers?

China leverages centralized planning, extensive renewable energy buildout, and ultra-high-voltage transmission infrastructure to deploy AI data centers at gigawatt scale, bypassing many US grid bottlenecks.

Will chip performance gains close the gigawatt power gap?

While chip efficiency improvements continue, the fundamental structural advantage in power infrastructure gives China an edge in deploying large-scale AI systems, making the power layer a critical bottleneck for the US.

What are the risks if the US cannot address its infrastructure constraints?

The US may face a ceiling on AI deployment capacity, potentially ceding technological and economic leadership in large-scale AI applications to China.

Source: ThorstenMeyerAI.com