Meta效仿SpaceX，将过剩AI算力变现

Meta has spent billions of dollars developing AI and building out data centers to support it. But now, the company may be preparing to put those data centers to a more immediately profitable purpose.

On Wednesday, Bloomberg reported that Meta is developing plans for a cloud infrastructure business, selling access to both AI compute power and models. The move would pit it against the big cloud providers like Amazon Web Services, Google Cloud, and Microsoft Azure. 

Meta’s decision to sell off excess compute comes weeks after SpaceX, via xAI, announced similar plans. In early May, SpaceX signed a deal with Anthropic to buy out all of the compute capacity at SpaceX’s Colossus 1 data center. SpaceX has signed similar leases since with Google and Reflection AI. The fact that Meta is doing the same is a signal that the winners of the AI race may not be the ones providing the best models and services, but rather the ones who own the data centers.

That is, if the demand for compute continues to hold, and if data centers retain their value. Some skeptics have warned the race to build out AI infrastructure is creating a bubble that leans heavily on rapidly depreciating chips. Others have questioned whether AI companies can generate enough end-user revenue to justify the trillion-dollar bets. 

Those concerns haven’t stopped Meta from investing heavily in infrastructure for AI compute. As of the end of the first quarter, Meta had committed to spending $182.9 billion on AI infrastructure in the coming years, including massive ongoing projects in Louisiana and Ohio. The Ohio project, which Zuckerberg said would be the size of Manhattan, is expected to come online this year.

Unlike Google and OpenAI, Meta hasn’t seen significant demand for its own AI models and services. Meta doesn’t break out its revenue from Meta AI or from Llama, its open-weight AI model family in its earnings, and executives have mostly emphasized the internal corporate uses of AI in public statements. That could mean that Meta’s AI endeavors don’t yet represent a material standalone revenue line. 

To get a return on some of its own colossal spend, Meta may copy CoreWeave’s business model and sell access to “raw” compute capacity, according to Bloomberg. The outlet also reported Meta is considering following AWS’s lead and selling access to various AI models — including its recently launched closed-weight model, Muse Spark — hosted on its AI infrastructure.

The new business line will be part of a new initiative reportedly dubbed Meta Compute, which is led by head of infrastructure Santosh Janardhan, Meta Superintelligence Labs leader Daniel Gross, and president Dina Powell McCormick.

The report confirms Zuckerberg’s May statements that a Meta cloud computing business is “definitely on the table” as a way to get a return on some of the massive investment into its strategy to develop AI “superintelligence.”