AI Compute Demand Surges Beyond Moore’s Law, Requiring $500 Billion Annually
The insatiable demand for AI compute has shattered the once-unassailable Moore’s Law, with OpenAI revealing that $500 billion must be invested annually until 2030 just to keep pace—exclusive of GPU or server costs. Data centers, now more expensive than U.S. office buildings, have become the new oil fields of the digital age, consuming capital at an unprecedented rate.
By 2028, global data center investment is projected to reach $900 billion, fueled by a 41% annual spike in AI server demand. The sector faces an $800 billion funding gap, requiring nearly $2 trillion to meet 2030 construction targets. U.S. data center construction costs alone have soared 322% since 2020, with $40 billion worth of facilities currently underway—a 400% increase from 2022.
