OpenAI is planning to build a dedicated data center as part of its Project Stargate, which aims to enhance its AI capabilities. The proposed data center could involve up to 5 exabytes of storage and will significantly increase the computing power available to OpenAI researchers, with plans to boost their capacity by eight times this year. The first phase of the Stargate infrastructure is expected to accommodate as many as 400,000 Nvidia AI chips, potentially making it one of the largest AI computing clusters. This development comes amid ongoing GPU shortages, which OpenAI CEO Sam Altman has highlighted, indicating that the company is currently out of GPUs but expects to receive tens of thousands soon, with hundreds of thousands more on the way. Additionally, there are speculations regarding a connection between OpenAI's data center plans and Microsoft's recent scaling down of its server infrastructure, which was initially intended to support OpenAI.