Oracle partners with AMD on 50,000-GPU AI supercluster

Oracle Cloud Infrastructure will deploy an AI supercluster powered by 50,000 AMD MI450 starting in the third quarter of 2026, with further expansion planned for 2027 and beyond.

“Demand for large-scale AI capacity is accelerating as next-generation AI models outgrow the limits of current AI clusters,” the companies said.

Oracle customers will gain access to AMD’s upcoming MI450 chips through Oracle Cloud Infrastructure offerings. The partnership aims to tap booming demand for infrastructure supporting tools like ChatGPT.

“By bringing together the latest AMD processor innovations with OCI’s secure, flexible platform and advanced networking powered by Oracle Acceleron, customers can push the boundaries with confidence,” said Mahesh Thiagarajan, executive vice president, Oracle Cloud Infrastructure.

Oracle’s expanded GPU options enable more flexibility for AI model training and inference. The deployment timeline aligns with Oracle’s partnership with OpenAI in which the latter is expected to buy $300 billion in computing power in about five years, according to the Wall Street Journal.