Meta is establishing Meta Compute, a group tasked with the ambitious growth of its computing infrastructure. Meta Compute intends to establish infrastructure that will use tens of gigawatts of energy within this decade and expand to hundreds of gigawatts in the future, Mark Zuckerberg revealed.

Meta Compute will be co-led by Santosh Janardhan, who is the head of global infrastructure and a co-head of engineering, along with Daniel Gross.

"Today, we are launching a new high-level initiative named Meta Compute," Zuckerberg stated in a post on Threads. Meta aims to generate tens of gigawatts during this decade and hundreds of gigawatts or beyond in the long run. The way we design, fund, and collaborate to create this infrastructure will serve as a competitive edge.

Janardhan's duties are both broad and highly technical: they will encompass Meta's complete system architecture, internal silicon initiatives, software framework, development tools, along with the establishment and daily management of the firm's global data center network and fleet.

Gross will lead an established team dedicated to long-term capacity planning and creating a supply chain that can provide equipment that will use gigawatts of power (i.e., a substantial number of chips and servers). His duties involve determining Meta's upcoming computing requirements, overseeing key supplier partnerships, tracking industry trends, and creating planning and business frameworks to facilitate infrastructure growth at a multi-gigawatt level.

Considering the duties of Janardhan and Gross, Meta is creating Meta Compute to methodically enhance AI-scale infrastructure at a scale well beyond conventional data-center expansion. With advanced AI models currently requiring computation in tens of gigawatts and hundreds of gigawatts over time, Meta must indeed obtain advanced hardware and build actual structures.

Also Read: Karnataka to Become Quantum Capital of Asia Soon

A committed organization enables Meta to prepare power, land, networking, and system architecture years ahead of time, instead of responding to increasing demand reactively.

The revised structure also consolidates control over the entire technical stack — encompassing software and system architecture, in-house silicon, networks, and data centers — to guarantee that hardware and software choices are made in tandem to enhance efficiency. Simultaneously, Meta Compute distinguishes between operational execution and the development of long-term capacity strategies and supply chain.

Also Read: Lookback 2025: 7 of the Year's Biggest Tech Acquisitions

Dina Powell McCormick will collaborate closely with the computing and infrastructure teams to guarantee that Meta's multi-billion-dollar investments are in line with the company's goals and provide real economic advantages in the areas where the company functions.

Also Read: Lookback 2025: 7 Indian Companies That Drove Major Expansions

Moreover, she will spearhead efforts to create new strategic capital partnerships and devise innovative methods to enhance Meta's long-term investment potential.