Zuckerberg Announces Meta Compute to Scale Long-Term AI Infrastructure

Meta on January 13 announced the launch of a new top-level initiative called Meta Compute to build a large-scale computing infrastructure for its long-term AI plans.

In a post on Facebook, CEO Mark Zuckerberg said the company plans to build “tens of gigawatts” of computing capacity this decade, with capacity expected to grow to “hundreds of gigawatts or more over time.” He added that the way Meta engineers, invests in, and partners to build this infrastructure will become a “strategic advantage”.

Meta said the new initiative will be led by Santosh Janardhan, head of global infrastructure and co-head of engineering at Meta, along with Daniel Gross, who is overseeing long-term AI capacity planning at the company.

Gross joined the company last year. He co-founded  Safe Superintelligence, alongside former OpenAI chief scientist Ilya Sutskever.

Janardhan will continue overseeing technical architecture, software systems, silicon efforts, developer productivity, and the company’s global data centre and network operations.

Gross will head a new group focused on long-term capacity planning, supplier partnerships, industry analysis, planning and business modelling.

Zuckerberg said the two leaders will work closely with former US deputy national security advisor Dina Powell McCormick, who recently joined Meta as the president and vice chairman. Her role includes engaging with governments and sovereign partners to help build, deploy, invest in and finance Meta’s infrastructure.

“I’m looking forward to working closely with Daniel, Santosh, Dina and their teams to scale Meta Compute and deliver personal superintelligence to billions of people around the world,” Zuckerberg wrote.

The announcement signals Meta’s intent to treat large-scale compute capacity as a core pillar of its future AI strategy, alongside software and model development.

The social media giant recently announced agreements linked to nuclear energy projects that could support up to 6.6 GW of new and existing electricity capacity in the United States by 2035, as the company plans for rising power demand from its data centres and AI infrastructure.

According to Meta, the agreements cover extended operations at existing nuclear plants, development of advanced nuclear reactors and long-term energy procurement. Meta said the projects will supply power to grids that support its operations, including its Prometheus AI supercluster in New Albany, Ohio.

Meta said the projects are expected to create thousands of construction jobs and hundreds of long-term operational roles, mainly in Ohio and Pennsylvania. The company added that it pays the full cost of energy used by its data centres, with the electricity delivered into regional grids.

The agreements include support for advanced nuclear developers TerraPower and Oklo, as well as long-term power purchases from operating plants owned by Vistra. Meta also referenced an earlier nuclear agreement signed with Constellation Energy last year.