The U.S. government will award GlobalFoundries $300 million to develop silicon photonics technology for more efficient AI data centers, a direct state subsidy to private capital. This funding arrives as major tech corporations like Amazon, Google, Microsoft, and Meta plan to spend an estimated $750 billion this year on AI infrastructure, nearly doubling last year's investment. This massive capital outlay fuels an AI expansion that drains global energy resources and concentrates computing power in the hands of a few.
Capital's Expansion
Hyperscalers are borrowing heavily to fund this artificial intelligence expansion. Reuters reported yields are rising as investor demand cools for this debt. The debt binge pushes yields up, making investors more selective about financing the buildout.
Amazon, Google, Microsoft, Meta, and Oracle are projected to spend about $750 billion this year on data centers, chips, and other AI infrastructure. This figure is up from roughly $400 billion last year, according to Goldman Sachs. By 2029, global AI infrastructure investment is forecast to top $1 trillion, a sharp increase from $318 billion last year, IDC reports.
Today, about 20 million AI chips are crammed into data centers worldwide, underpinning the technology’s growing abilities. This figure is expected to double roughly every nine months. The world is on pace to have about 200 million of these chips by the end of 2028, ten times current levels.
In March 2024, the world had about 2.4 million “H100 equivalent” AI chips. Chinese companies held roughly 1.16 million H100-equivalent chips by the end of 2025, up from about 244,000 at the beginning of 2024, though these figures exclude smuggled chips and other offshore computing resources used by Chinese firms, Epoch AI estimated.
The State's Hand
The U.S. government's $300 million award to GlobalFoundries aims to develop silicon photonics technology for faster AI chip links. This direct public funding supports private sector innovation and profit.
The United States controls about 80% of global computing power that drives AI, according to Epoch AI. It is home to approximately 5,500 data centers, ten times more than the next closest country. U.S. companies like Amazon, Google, Microsoft, and Meta control about 80 percent of this global computing power.
This dominance is part of a broader U.S.-China race for technological supremacy. This competition threatens to leave the rest of the world behind.
European Union nations, including France and Germany, are attempting to encourage data center construction. The EU currently holds only 5 percent of global AI computing power. Europe's efforts are hampered by issues like electricity and land access, permitting, and financing, revealing the limits of national capital in a globalized struggle.
The Persian Gulf has pledged billions to build data centers. However, the ongoing war in Iran has affected these plans, demonstrating how geopolitical conflicts serve to disrupt or redirect capital flows.
China’s top tech companies are also building new AI chips and data centers. Huawei, ByteDance, and Alibaba are expected to spend $111 billion this year on data centers and other AI investments, Bernstein Research reports. China’s National Energy Administration estimates the country’s electricity use for data centers will reach around 91 gigawatts by 2030, about 6 percent of total use, up from 19 gigawatts last year.
Resource Extraction
Last year, data centers consumed 64 gigawatts of electricity globally. This is roughly as much electricity as Germany consumes.
By the end of 2030, this consumption is expected to quadruple. It will eclipse the power used by all countries in South America and Africa combined.
At the most advanced AI data centers, every gigawatt of power equates to roughly $40 billion to $60 billion in costs. These costs include servers, land, connectivity, and utility hookups, according to industry estimates. This massive expenditure represents a significant drain on collective energy resources for private profit.