Meta Platforms plans to invest billions of dollars on AI compute infrastructure, which includes building colossal data centre clusters.
In a Facebook post, CEO Mark Zuckerberg outlined plans to build several multi-gigawatt super clusters with the first expected to come online next year. These facilities will be used to train and manage workloads generated by Meta’s advanced AI models.
The social media giant retooled its AI efforts last month under a newly formed division called Meta Superintelligence Labs. Zuckerberg selected former Scale AI CEO Alexandr Wang to lead the MSL team. Wang, who is also chief AI officer for Meta, was CEO of Scale AI prior to Meta Platforms making a huge investment in the company last month.
“Meta Superintelligence Labs will have industry-leading levels of compute and by far the greatest compute per researcher”, Zuckerberg stated in the post. He added the company has the capital needed to build the data centre clusters, the first of which is called Prometheus. The second cluster, Hyperion, is intended to be able to scale up to five gigawatts over several years. “Just one of these covers a significant part of the footprint of Manhattan”, he said.
Source: Mobile World Live
Image Credit: Meta
