Stop overpaying - start transferring money with Ogvio. Sign up, invite friends & grab Rewards now! 🎁
OpenAI Supercharges Stargate as Elon Musk Targets 50 Million AI Chips
Key Takeaways
- OpenAI and Oracle will add 4.5 gigawatts to Stargate, which will push the project past 5 gigawatts total;
- Sam Altman expects 1 million GPUs online by 2025 and says OpenAI may exceed its $500 billion plan;
- Elon Musk aims for xAI to hit 50 million H100-equivalent chips in 5 years.
OpenAI has partnered with Oracle to boost its artificial intelligence (AI) computing capacity by an additional 4.5 gigawatts (GW).
The deal, published on July 22, expands the existing Stargate I site in Abilene, Texas. In January, OpenAI shared plans to roll out 10 GW of AI infrastructure across the US, but this new step pushes that target even further.
OpenAI CEO Sam Altman shared the news on X, and called it a "gigantic infrastructure project". He added that over one million GPUs are expected to be online by the end of 2025.
Did you know?
Subscribe - We publish new crypto explainer videos every week!
What is Defi 2.0? (Explained with Animations)
This new agreement brings the Stargate project’s total development pipeline to more than 5 GW. That amount of power could support over two million AI chips.
Altman mentioned that OpenAI is planning to exceed its original $500 billion commitment for Stargate.
Following OpenAI’s announcement, Elon Musk revealed that xAI is aiming to reach the equivalent of 50 million H100 AI chips within five years. According to a post by another X user, that would be 500 times more computing power than last year’s most powerful AI supercomputer.
xAI’s upcoming Colossus 2 system is expected to run on 550,000 GB200 chips, which together match about 5.5 million H100s.
Additionally, xAI recently signed a $200 million agreement with the US Department of Defense. What does the deal cover? Read the full story.