OpenAI agreed a multi-year compute partnership with Amazon Web Services reported to secure very large Nvidia-GPU capacity across AWS data centres. The arrangement diversifies OpenAI’s cloud footprint away from single-vendor reliance and aims to support large-scale model training and inference. Market reaction lifted Amazon’s cloud narrative as investors priced stronger long-term AI infrastructure demand. Analysts note the deal may accelerate enterprise AI rollouts but warn it also concentrates GPU demand and could pressure long-term pricing and supply across hyperscale providers and chip vendors.