Photo by cottonbro studio on Pexels

In a strategic move to bolster its AI compute infrastructure, OpenAI is diversifying its cloud partnerships, moving beyond its previous exclusive reliance on Microsoft. The company has forged new agreements with Amazon Web Services (AWS) and Oracle, supplementing its ongoing relationship with Microsoft.

This multi-cloud strategy underscores the increasing demand for high-performance GPUs and the necessity of substantial, long-term capital investments to secure the required AI compute resources. The deal with AWS provides OpenAI access to a wide array of NVIDIA GPUs and CPUs, enabling it to support both the intensive model training and the real-time inference demands of applications like ChatGPT.

Industry analysts suggest that this trend towards multi-cloud deployments is driving other organizations to explore managed platforms such as Amazon Bedrock, Google Vertex AI, and IBM watsonx, where cloud providers manage the underlying infrastructure and associated risks. The move by OpenAI also highlights the potential risks of relying solely on a single cloud provider for AI workloads, emphasizing the importance of strategic AI budgeting as a core element of corporate capital planning.