Meta Capital Spending Hits $8.5B, Led by AI Servers and Data Centers: Will Grow ‘Significantly’ in 2025
While Mark Zuckerberg says Llama 4 will require 10 times more processing than its predecessor Meta has continued to spend heavily on servers, data centers, and network infrastructure as it builds out its AI infrastructure. Capital spending hit $8.5 billion for the quarter, and the company warned that it expects costs to rise further. Facebook’s […]
While Mark Zuckerberg says Llama 4 will require 10 times more processing than its predecessor
Meta has continued to spend heavily on servers, data centers, and network infrastructure as it builds out its AI infrastructure.
Capital spending hit $8.5 billion for the quarter, and the company warned that it expects costs to rise further. Facebook’s parent company reported revenue of $39.07 billion, up 22% from last year and beating analysts’ expectations. It also provided a revenue forecast of between $38.5 billion and $41 billion for the next quarter. Net income was $13.46 billion, up 73%.
Capital spending fell short of analysts’ expectations of $9.51 billion, but still grew 32.8% year-over-year. Chief Financial Officer Susan Li said the company is “clearly on the path to building significant capacity.”
She added: “We are adopting a strategy of organizing our data centers in multiple stages of development, which allows us to adapt to meet increased demand and shorter lead times if necessary, while also limiting the spending that we commit to in the coming years.”
“So while we plan to significantly increase capital spending in 2025, we believe we have a good framework in terms of thinking about where the opportunities are and making sure we have the flexibility to implement them in the way that makes the most sense.”
The company will use the same infrastructure it built for generative AI training for generative AI inference. “We can also use it to make rankings and recommendations, making some adjustments, like adding general computation and storage,” Li said.
“There are many use cases for the life of each individual data center.”
While costs will increase next year, Li said the exact level of spending is still being determined. “We’re working across a broad range of capital expenditure outcomes right now based on different data center options in different locations,” he said. “It’s really a very dynamic area of planning for us right now, and we’ll have more clarity on that in the fourth-quarter call.”
Despite the spending increase, generative AI “is not expected to be a significant revenue driver in 2024,” Meta said. However, CEO Mark Zuckerberg said he would “rather risk building capacity before it’s needed, rather than too late, given the long lead times to get new infrastructure projects up and running.”
Giving a sense of the scale needed to keep up with generative AI, he said: “The amount of compute needed to train [the AI model] Llama 4 will likely be close to 10 times what we used to train Llama 3, and future models will continue to grow beyond that.”
https://www.datacenterdynamics.com/en/news/meta-capex-hits-85bn-driven-by-ai-servers-and-data-centers-will-grow-significantly-in-2025/
Was this article useful to you?
0 Feedbacks