OpenAi expands Stargate with Oracle to surpass 5 GW of AI energy, while Elon Musk describes plans for Xai deploy 50 million H100 scale units in 5 years.
OpenAi has announced an expansion of 4.5 Gigawatt in partnership with Oracle to boost AI’s future development.
The agreement, part of OpenAi’s long-term vision to deploy 10 US computing gigawatts, will increase its existing Stargate I installation in Abilene, Texas, and push the project beyond its original White House commitment in January, the company said on Tuesday.
“This is a gigantic infrastructure project,” published OpenAi Sam Altman CEO in X, sharing images from the Abilene website. In a previous post, he confirmed that more than 1 million GPUs will be online by the end of the year and joked: “Now they better work out how 100x this.”
Oracle’s additional agreement will bring Stargate’s total development pipeline to more than 5 GW, enough to feed more than 2 million AI chips. “We are planning to significantly expand Stargate’s ambitions after the $ 500 billion commitment we announced in January,” added Altman.
Musk reveals the bold plan for Xai
After the announcement of OpenAi, Elon Musk also shared a bold plan for his Iai company. “@Xai’s goal is 50 million in equivalent-equivalent computing units H100 (but much better power efficiency) online within 5 years,” wrote the billionaire in a Tuesday post at X.
According to estimates of the X user Teslaprice, this would represent 500 times the computer power of what was considered the most powerful AI supercomputer in the world for just a year.
The Xai Supercomputer Colossus 2, set to air soon, will use 550,000 GB200 chips, approximately 5.5 million H100s. Musk’s plan with Xai, if done, almost 10x.
“Elon is saying that they will reach the equivalent of 50 million H100 in 5 years. So this will be 500x of the state -of -the -art cluster for 12 months,” the user estimated.
$ 500b Stargate Ai Project faces challenges
Earlier this year, US President Donald Trump announced the launch of Stargate, a US $ 500 billion AI infrastructure initiative led by the private sector. The project is supported by OpenAi, Softbank and Oracle, with the aim of creating AI data centers in the US and creating over 100,000 jobs.
However, according to a recent report from the Wall Street Journal, the initiative faced major delays and internal disagreements between the main partners of Softbank and OpenAi. Despite the initial promises of deploying $ 100 billion immediately, the project reduced its short -term goals to build a single data center by the end of the year, according to the report.