Chasing the AI Wave: Meta Invests Tens of Billions to Build a World-Class Computing Power Network

Glonghui, January 13th｜Meta CEO Mark Zuckerberg announced via Threads on Monday that the company will launch a top-level strategic initiative called Meta Compute. The plan aims to build computing infrastructure on the scale of tens of gigawatts (GW) within this decade, with intentions to further expand to hundreds of gigawatts in the future. According to Zuckerberg’s post on social media, the Meta Compute initiative will be co-led by Santosh Janardhan, Meta’s Global Head of Infrastructure and Co-Head of Engineering, along with Daniel Gross. Zuckerberg stated that they will work closely with Dina Powell McCormick, who recently joined Meta as President and Vice Chair. Janardhan will continue to oversee Meta’s technical architecture, software stack, chip programs, development efficiency, as well as the construction and operation of the company’s global data center clusters and networks. Gross will lead a newly formed group focused on long-term capacity strategy, vendor partnerships, industry analysis, planning, and business modeling. After the lukewarm market response to the Llama 4 model, Meta has been striving to maintain its standing in the Silicon Valley AI race and has committed to a capital expenditure of up to $72 billion in 2025.