Prometheus, Hyperion, and Titan: These are the mega-constructions Meta is planning to win the AI race.

Seeking to position itself ahead of OpenAI and Google in the artificial intelligence race, Mark Zuckerberg has announced via Threads that Meta is investing "billions of dollars" in building several data centers to boost its AI capabilities.
The company is focused on advancing the development of "superintelligence" , for which, at the beginning of July, it revealed the creation of its new Superintelligence Labs team , led by the former CEO of the startup Scale AI, Alexandr Wang, and the former CEO of GitHub, Nat Friedman. Furthermore, with the construction of these infrastructures, made up of multiple data centers and computing clusters located in different parts of the United States , Meta is willing to promote its strategy to mark a before and after in the sector.
However, through new infrastructure megaprojects called Hyperion and Prometheus, as well as the new Titan clusters, Meta is not only seeking to increase its capacity to train the most powerful artificial intelligence models, but also to transform its capacity for innovation in the industry.
This is what Meta's new data centers look like.The names of Meta's new data centers are inspired by the Titans of Greek mythology, symbolizing the scale and significance of these megaprojects.
Prometheus will be the company's first data center with a capacity greater than 1 gigawatt , will be located in Ohio and is expected to be operational in early 2026, becoming one of the first data centers of this size and capacity directly managed by a technology company.
Separately, according to Mark Zuckerberg's Threads post, the company is also preparing to begin construction of a data center called Hyperion . This infrastructure will be located in Louisiana, will offer increased capacity because it will be able to scale up to five GW "in several years," and will have an area equivalent to "a significant portion of the surface area of Manhattan."
Finally, Meta will also develop multiple clusters called Titan . These computing infrastructures will process the workloads of future artificial intelligence models, and their location is currently unknown.
However, Zuckerberg states in the Threads post that the development of these infrastructures is intended to boost Meta's artificial intelligence capabilities , so the new data centers will feature "industry-leading levels of computing."
A multi-million dollar investmentMeta's new data centers are run by Meta Superintelligence Labs , the division created by the company to focus all its efforts and research on advanced artificial intelligence.
According to Zuckerberg, these infrastructures will allow Meta Superintelligence Labs teams to operate with industry-leading computing capabilities, considering that their implementation could be part of a talent attraction strategy , with which Meta would seek to attract the best AI researchers.
However, the scale and investment required for these projects has raised some concerns among analysts and investors. Therefore, in the face of this, Zuckerberg defends the financial viability of the initiative, ensuring that Meta has "the necessary operating capital."
Sign up for our newsletter and get the latest technology news straight to your inbox.
20minutos