OpenAI publishes clarification on Oracle and Microsoft partnerships

Generative AI firm OpenAI said its strategic cloud partnership with Microsoft remained unchanged following a strategic agreement with Oracle.

“We want to clarify some inadvertent confusion. Our strategy cloud relationship with Microsoft is unchanged,” wrote OpenAI on X on Thursday.

This statement follows a team-up between OpenAI and Oracle to use Oracle Cloud Infrastructure (OCI) to extend the Azure AI platform. OCI is designed with high-performance NVIDIA GPU instances to run workloads including generative AI, computer vision, natural language processing, recommendation systems, and other applications.

OpenAI says OCI enables the use of the Azure AI platform on OCI infrastructure for inference and other requirements. Meanwhile, all pre-training of frontier models will continue on computing systems built on Microsoft cloud infrastructure.

Last year, OpenAI and Microsoft announced a long-term collaboration through a multiyear, multibillion-dollar investment in AI development. This agreement extends beyond 2019 and 2021. It covers specialized supercomputing systems based on the Azure AI platform.

The agreement includes Microsoft Azure OpenAI Service for AI applications through direct access to OpenAI models using Azure enterprise’s capabilities and AI infrastructure and tools. Interestingly, Azure became an exclusive cloud provider to run all OpenAI workloads across research, products, and API services.

So, the new partnership underlines the last paragraph which may raise questions from the online community on extending the Azure infrastructure with OCI.