On January 13, Meta announced the launch of a new top-level initiative called Meta Compute to build large-scale computing infrastructure for long-term AI plans.
CEO Mark Zuckerberg said in a Facebook post that the company plans to build “tens of gigawatts” of computing power over the next decade, with that capacity expected to increase to “hundreds of gigawatts or more over time.” He added that the way Meta engineers, invests and partners to build this infrastructure will be a “strategic advantage.”
Meta said the new effort will be led by Santosh Janardhan, Meta's head of global infrastructure and co-head of engineering, and Daniel Gross, who oversees the company's long-term AI capability planning.
Gross joined the company last year. He co-founded Safe Superintelligence with former OpenAI chief scientist Ilya Sutskever.
Mr. Janardhan will continue to oversee technology architecture, software systems, silicon efforts, developer productivity, and operations of the company's global data centers and networks.
Mr. Gross will lead a new group focused on long-term capacity planning, supplier partnerships, industry analysis, planning and business modeling.
Zuckerberg said the two leaders will work closely with former deputy national security adviser Dina Powell McCormick, who recently joined Meta as president and vice chair. Her role will include working with governments and sovereign partners to help build, deploy, invest and finance Meta's infrastructure.
“I look forward to working closely with Daniel, Santosh, Dina, and their team to scale metacomputing and bring personal superintelligence to billions of people around the world,” Zuckerberg wrote.
This announcement signals Meta's intention to treat large-scale computing power as a core part of its future AI strategy, alongside software and model development.
The social media giant recently announced deals related to nuclear power projects that could support up to 6.6 gigawatts of new and existing power capacity in the US by 2035, as the social media giant plans to increase power demand from data centers and AI infrastructure.
Mehta said the agreement includes extending the operation of existing nuclear power plants, developing advanced reactors and long-term energy procurement. Meta said these projects will provide power to the electrical grid that supports its operations, including the Prometheus AI Supercluster in New Albany, Ohio.
Mehta said the projects are expected to create thousands of construction jobs and hundreds of long-term operational roles, primarily in Ohio and Pennsylvania. The company added that it fully pays for the energy used in its data centers by supplying power to the local power grid.
The agreement includes support for advanced nuclear developers TerraPower and Oklo, as well as long-term power purchases from operating power plants owned by Vistra. Mehta also pointed to an earlier nuclear agreement signed last year with Constellation Energy.
