Decentralized AI Training Can Cut Cooling and Carbon, but the Network Bill Still Keeps Frontier Models Centralized
Decentralized AI training is not a simple replacement for giant GPU clusters. Its real advantage is narrower: spreading workloads across locations can reduce cooling demand and make cleaner electricity easier to use, but once training depends on tight coordination across many sites, bandwidth, latency, and fiber costs start eating away at those gains. The energy…