Post by : Bianca
Haleem
OpenAI has agreed to a multi-year partnership valued at $38 billion with Amazon Web Services, securing broad access to AWS’s high-performance cloud infrastructure to scale demanding AI workloads.
The arrangement follows OpenAI’s recent corporate restructuring that removed its non-profit framework, opening the company to wider commercial opportunities. The move also broadens cloud provider relationships beyond Microsoft.
Under a seven-year deal, OpenAI will run on AWS EC2 UltraServers equipped with thousands of Nvidia GB200 and GB300 GPUs. The setup is intended to support inference for current ChatGPT models and the training of future systems, with the ability to scale across millions of CPUs. AWS has highlighted the configuration’s efficiency, performance and security, with full deployment expected by the end of 2026 and potential extension into 2027.
OpenAI has been expanding its cloud footprint: its foundation models were made available on Amazon Bedrock earlier this year, and the company maintains separate compute arrangements with Microsoft Azure, Google Cloud and Oracle to ensure sufficient capacity.
The partnership reflects a strategic shift toward diversifying infrastructure partners. By leveraging AWS’s capabilities, OpenAI aims to accelerate research and commercial rollouts while reducing dependence on a single cloud provider amid intensifying competition for large-scale compute resources.
English
