OpenAI and Oracle have reached a cloud services agreement worth up to $30 billion, marking one of the largest cloud computing collaborations in the field of artificial intelligence (AI) to date. This partnership will also advance OpenAI’s “Stargate” data center project.
The news has had a positive impact on Oracle’s stock, with shares rising by 3.9% to reach $227.46 per share, hitting a historic high. The stock has seen a cumulative increase of 36% so far this year.
In January, President Trump met with several high-tech company executives at the White House, emphasizing the importance of the “Stargate” project. This AI infrastructure initiative, jointly promoted by OpenAI, SoftBank from Japan, and Oracle, is expected to create over 100,000 job opportunities in the United States.
Earlier this week, Oracle announced a cloud computing contract expected to generate $30 billion in annual revenue starting in 2028. This deal’s annual revenue is roughly three times Oracle’s $10.3 billion infrastructure business income for 2025.
According to sources familiar with the matter who spoke to the Financial Times and Bloomberg, the customer for this contract is OpenAI.
It was revealed that in order to implement the “Stargate” project, OpenAI will lease a total of about 4.5 gigawatts (GW) of data center computing capacity from Oracle. This capacity is equivalent to a quarter of the total operating capacity of data centers in the United States.
Oracle plans to expand its power capacity from the current 1.2GW to around 2GW at its existing data center in Abilene, Texas. Additionally, they are considering constructing new data centers in states like Michigan, Wisconsin, Wyoming, New Mexico, Georgia, Ohio, and Pennsylvania.
Typically, 1GW of power can supply the electricity needs of approximately 750,000 households, and a 4.5GW capacity is equivalent to the power output of several nuclear reactors, effectively providing power to millions of households.
OpenAI and its partners are aggressively expanding the “Stargate” project, with a total investment expected to reach $500 billion over the next four years.
As per previous announcements by OpenAI, the initial investment amount is $100 billion, with some of it used to acquire 875 acres of land in Abilene, Texas, planning to build 10 data centers and deploy a total of 700,000 GPUs.
Internal forecasts presented by OpenAI to investors indicate that by 2030, the data centers of Stargate will support three-quarters of the computational power needed to run and develop AI models for OpenAI. This suggests that within the next five years, OpenAI will gradually reduce its dependence on Microsoft.
To meet the massive computational power demands of the “Stargate” project, Oracle plans to purchase approximately 400,000 new Nvidia GB200 chips, with a total cost of up to $40 billion, specifically for the data center in Abilene, Texas.
Aside from the United States, the “Stargate” project is also actively expanding globally.
In May of this year, OpenAI announced a partnership with UAE-based AI company G42 to build the world’s largest AI super data center in the region. The data center is expected to occupy approximately 10 square miles, with a total power demand of 5GW. OpenAI will be the main tenant of this data center, utilizing 1GW of the computational resources. This project is supported by the UAE Sovereign Wealth Fund and receives technical assistance from Nvidia and Cisco Systems.