What happened
Brookings published a projection on Wednesday putting cumulative AI infrastructure investment at $10. 3 trillion by 2032, according to a summary from Crypto Briefing. The scope covers hyperscale data centers, dedicated power generation, high-density networking gear, and custom silicon from vendors like Nvidia, AMD, and Broadcom.
Brookings frames the spend as a structural rebuild of the compute stack rather than a cyclical capex bump. The report follows a run of similar forecasts from McKinsey and Goldman Sachs earlier this year, all clustered in the $7T to $11T range through the early 2030s. The direction is consistent.
The order of magnitude is now the consensus.
Why it matters
A $10. 3 trillion capital wave doesn't stay in one sector. Grid operators from ERCOT to PJM are already flagging AI data center demand as the binding constraint on new interconnects, and the same power shortage that squeezes hyperscalers also squeezes bitcoin miners.
Chip supply is the other pinch point. Every H100 or MI300X that ships to a hyperscaler is one that doesn't ship to a mining ASIC foundry line or a decentralized GPU network. For crypto, the read-through is that AI infrastructure isn't a parallel story.
It's the same physical stack, competing for the same watts and wafers.
Market impact
There was no direct token reaction to the Brookings report itself, and no coins are named in the underlying data. The market has priced AI capex expectations for months. What the report does is harden the base case.