OpenAI’s Trillion-Dollar AI Infrastructure Plan is Kicking Off an Investment Gold Rush

“`html

OpenAI’s “Stargate” Project Overview

Metric	Details
Total Investment	$500 billion USD over 4 years
Oracle Lease	$30 billion USD annually for 4.5 billion watts of computing
First Phase (Stargate UAE) Go-Live	2026
Stargate UAE Initial Chip Capacity	Up to 100,000 Nvidia chips

OpenAI’s AI infrastructure plans are no longer just talk — CEO Sam Altman is gearing up to pour trillions of dollars into the hardware needed to power the future of artificial intelligence.

This isn’t a distant dream; the first massive chess move is already on the board.

In July, OpenAI locked in a deal to lease a staggering 4.5 billion watts of computing power from Oracle (Jiaguwen 甲骨文).

The price tag? Around $30 billion USD annually.

This is easily one of the largest cloud service agreements the AI world has ever seen.

And it’s all fueling OpenAI’s ridiculously ambitious “Stargate” data center project.

The “Stargate” Project: A $500 Billion Supercluster Takes Shape

Let’s be clear: Stargate is a beast of a project.

The plan is to drop a cool $500 billion USD over the next four years to build out new AI infrastructure for OpenAI, both in the United States and across the globe.

The goal is simple: acquire enough raw computing power to build the next generation of AI models and satisfy the insane consumer demand for tools like ChatGPT.

The first piece of this global puzzle is already being assembled.

Back in May, Nvidia (Yingweida 英伟达) announced a partnership with a heavyweight crew including G42, OpenAI, Oracle (Jiaguwen 甲骨文), SoftBank Group, and Cisco (Sike 思科).

Together, they’re launching “Stargate UAE,” a next-gen AI infrastructure cluster in Abu Dhabi.

Here’s the breakdown:

• The first phase is slated to go live in 2026.
• It could be armed with as many as 100,000 Nvidia chips right out of the gate.

Analysts at Cinda Securities (Xinda Zhengquan 信达证券) see this as confirmation of the AI infrastructure build-out everyone’s been predicting.

As Stargate keeps moving, investment will only get bigger, pumping a steady stream of orders and capital into the AI computing supply chain.

A Global Spending Spree: The AI CAPEX Tsunami

It’s not just OpenAI. The entire tech world is in an AI infrastructure arms race.

Marvell, a key semiconductor player, predicts North American Cloud Service Providers (CSPs) are on track to spend $595 billion USD in 2025 alone.

By 2028, that number is expected to hit $1.022 trillion USD. That’s a 20% Compound Annual Growth Rate (CAGR). Wild.

Research from Guosen Securities (Guoxin Zhengquan 国信证券) shows a tidal wave of government and corporate cash flooding into AI infrastructure since the start of 2025:

• China: China Construction Bank (Zhongguo Yinhang 中国银行) has committed over ¥1 trillion RMB ($140 billion USD).
• United States: An initial $100 billion USD is allocated, set to expand to $500 billion USD within four years.
• European Union: A massive €200 billion (approx. $210 billion USD) is earmarked.
• France: The country has pledged €109 billion (approx. $114 billion USD).

Projected Cloud Service Provider Spending (North America)

Year	Projected Spending (USD)
2025	$595 billion
2028	$1.022 trillion

Global Government & Corporate AI Infrastructure Commitments (from 2025)

Region	Commitment (Approx. USD)
China	$140 billion (¥1 trillion RMB)
United States	$100 billion (initial), expanding to $500 billion
European Union	$210 billion (€200 billion)
France	$114 billion (€109 billion)

Big Tech Opens the Floodgates

The tech giants are leading the charge, projecting a combined capital expenditure (CAPEX) of $320 billion USD for fiscal year 2025 — a 39% jump from 2024.

And Chinese tech titans are matching the pace:

• ByteDance (Zijie Tiaodong 字节跳动): Plans to spend over ¥150 billion RMB ($20.64 billion USD) in 2025, mostly on AI.
• Alibaba (Ali 阿里): Its Q1 2025 CAPEX hit ¥24.6 billion RMB (approx. $3.45 billion USD), a 121% year-over-year increase. CEO Wu Yongming (Wu Yongming 吴泳铭) plans to invest over ¥380 billion RMB (approx. $53.3 billion USD) in cloud and AI over the next three years. That’s more than they spent in the entire last decade.
• Tencent (Tengxun 腾讯): Tencent’s CAPEX has been explosive. Q4 2024 saw a 386% year-over-year jump to ¥36.6 billion RMB (approx. $5.14 billion USD). Q1 2025 continued the trend with a 91% YoY increase to ¥27.5 billion RMB (approx. $3.86 billion USD).

Selected Big Tech AI-Related CAPEX

Company	CAPEX Details
Combined Big Tech	$320 billion USD for FY2025 (39% jump from 2024)
ByteDance	$20.64 billion USD (¥150 billion RMB) planned for 2025 (mostly AI)
Alibaba	Q1 2025 CAPEX: $3.45 billion USD (¥24.6 billion RMB), 121% YoY increase. Plans over $53.3 billion USD (¥380 billion RMB) in cloud/AI over 3 years.
Tencent	Q4 2024 CAPEX: $5.14 billion USD (¥36.6 billion RMB), 386% YoY. Q1 2025 CAPEX: $3.86 billion USD (¥27.5 billion RMB), 91% YoY.

Who Wins? The Benefiting Sectors in the AI Supply Chain

So, where is all this money going?

The AI infrastructure supply chain is basically split into two parts:

1. General Infrastructure: Think building construction, cooling systems, and power grids.
2. IT Infrastructure: This is the sexy stuff — AI chips, servers, network gear, and storage devices.

Guosen Securities (Guoxin Zhengquan 国信证券) points out that AI Data Centers (AIDC) are the core focus of this spending boom.

As CAPEX hits an inflection point, the industry is kicking off a long-term boom cycle. Two areas, in particular, are set to benefit massively due to their high value and technical importance: power supply systems and cooling systems.

Keep an eye on companies that are core suppliers and are already expanding with the big players.

Key Areas and Companies to Watch:

1. Power & Gas Turbines: Specifically, the makers of hot-end blades and other critical castings.
   • Yingliu Shares (Yingliu Gufen 应流股份)
   • Haomai Technology (Haomai Keji 豪迈科技)
2. Cooling Systems: Especially liquid cooling, chiller units, and the compressor manufacturers that supply them.
   • Binglun Environment (Binglun Huanjing 冰轮环境)
   • Hanchung Precision Machinery (Hanchung Jingji 汉钟精机)
   • Liande Shares (Liande Gufen 联德股份)
   • Tongfei Shares (Tongfei Gufen 同飞股份)

The “Golden Age” of Liquid Cooling

The market is already catching on. Liquid-cooled server stocks have been on an absolute tear.

Check out these recent gains:

• Dayuan Pump Industry (Dayuan Bengye 大元泵业): Surged over 98% in half a month.
• Invic (Yingweike 英维克): Jumped 93.45%.
• Oulutong (Oulutong 欧陆通): Up 82.52% in just five days.
• Feilong Corporation (Feilong Gufen 飞龙股份): Gained over 66%.

• Dayuan Pump Industry (Dayuan Bengye 大元泵业): Surged over 98% in half a month
• Invic (Yingweike 英维克): Jumped 93.45%
• Oulutong (Oulutong 欧陆通): Up 82.52% in just five days
• Feilong Corporation (Feilong Gufen 飞龙股份): Gained over 66%

Why the frenzy? A research report from Minsheng Securities (Minsheng Zhengquan 民生证券) nails it: “power” is the biggest bottleneck holding back AI.

As AI chips get more powerful, they generate insane amounts of heat, which throttles performance.

Liquid cooling is seen as the “panacea” for this problem.

The takeaway is clear: as AI development accelerates, the data center liquid cooling industry is entering its “golden age.”

The massive investment in OpenAI’s AI infrastructure and the broader tech industry is creating a once-in-a-generation opportunity for the companies building the picks and shovels for this digital gold rush.

References

• OpenAI Plans Trillion-Dollar AI Infrastructure; Institutions Foresee Funding Inflection Point, Deeply Benefiting Key Sectors – Eastmoney Finance
• OpenAI Official Website
• Oracle Official Website
• Nvidia Newsroom

“`