Meta has formed a new initiative called “Meta Compute” to oversee the planning, deployment, and operations of its growing fleet of AI datacenters.

“Meta is planning to build tens of gigawatts this decade, and hundreds of gigawatts or more over time. How we engineer, invest, and partner to build this infrastructure will become a strategic advantage,” wrote CEO Mark Zuckerberg in a post to his Twitter clone, Threads, on Monday.

The news comes just hours after Zuckerberg revealed that Dina Powell McCormick, who previously spent 16 years at Goldman Sachs and served as an advisor to President Trump, had joined the company as President and Vice Chairman.

According to Zuck, Meta Compute will be led by Santosh Janardhan, the company’s head of global infrastructure, and Daniel Gross, who joined Meta’s Superintelligence team in mid-2025.

“Santosh will continue to lead our technical architecture, software stack, silicon program, developer productivity, and building and operating our global datacenter fleet and network,” he explained. “Daniel will lead a new group responsible for long-term capacity strategy, supplier partnerships, industry analysis, planning, and business modeling.”

Both executives will work closely with McCormick, who Zuck noted would be focusing much of her energy on “partnering with governments and sovereigns to build, deploy, invest in, and finance Meta’s infrastructure.”

The initiative comes as Meta continues to spend lavishly in pursuit of delivering “personal superintelligence” to the masses. Meta forecast $72 billion capital expenditure in pursuit of that goal during its fiscal 2025, and higher spending in the following year.

However, compared to its competitors, Meta doesn’t have much to show for all of its spending. Following the lackluster reception of its open-source Llama 4 models, a talent war with OpenAI, and the departure of machine learning guru Yann Lecun, the company’s next-gen foundation models remain shrouded in mystery.

After touting open source model development, Zuckerberg has reportedly abandoned Llama and pivoted to building proprietary models under the codenames Avocado and Mango. However, at the same time, Meta continues to release models like its Segment Anything Series (SAM) in the open.

Despite missing the mark on Llama 4, Zuckerberg clearly hasn’t given up on his generative AI aspirations and is pushing ahead with ever more grandiose infrastructure projects.

Meta is currently working on multiple gigawatt-scale datacenter construction projects across Ohio, Louisiana, and Texas, plus other locations.

To keep this expansion from running out of steam, Meta last week signed three new long-term contracts with TerraPower, Oklo, and Vistra for nuclear energy. Combined with the company’s existing commitments with Constellation Energy, the Social Network has now contracted for roughly 6.6 gigawatts of atomic power. ®