OpenAI is projected to spend over $852 billion by 2030, accounting for $748 billion in remaining performance obligations (RPOs) for Microsoft, Amazon, and Oracle. Additional RPOs totaling at least $70 billion involve Cerebras, CoreWeave, Nebius, IREN, Lambda, and Nscale, alongside planned spending on Broadcom “Jalapeno” chips. The entity intends to spend $50 billion or more on compute this year, representing over 50% of global AI compute infrastructure spend.