OpenAI and Amazon Web Services (AWS) have struck a monumental $38 billion deal to supply computing power via Nvidia chips over the next seven years, marking one of the largest infrastructure agreements in AI history. The contract underscores OpenAI’s need for scalable high-performance hardware as model sizes and compute demand soar. Experts say this fuels an arms-race in cloud AI infrastructure and signals a premium on AI firms securing long-term chip supply and capacity. The deal also highlights AWS’s ambition to lock in hyperscale AI workloads.