Source: OpenAI, a major Nvidia customer, plans to buy Google's TPUs; a job listing shows it seeks an engineer to "lead the optimization" of its models on TPUs (Ryan Barwick/The Information)

Oracle Cloud Infrastructure (OCI) customers will soon have a more diverse selection of AI resources. During a recent event in Las Vegas, Oracle expanded its partnership with OpenAI and plans to bring Google's Tensor Processing Units (TPUs) to OCI. This move will make OCI the first hyperscale cloud platform outside Google's own cloud services to offer TPUs.

OCI president Mahesh Thiagarajan explained that different AI workloads have varying compute needs, and adding Google TPUs to OCI will give customers the flexibility to choose the right infrastructure for each workload's performance, throughput, and economics. Despite this move away from Nvidia, Oracle's partnership with OpenAI continues, with new ChatGPT integrations announced alongside two products that incorporate OpenAI models.

The first product, Oracle AI Workspace, is designed to make Oracle's existing business platforms agent-ready and accessible through ChatGPT plugins. The second product, Oracle AI Security Mesh, is a managed platform that combines frontier AI models for cybersecurity with Oracle's security expertise. Additionally, Oracle announced a deal with Snowflake to bring its AI data cloud to OCI, allowing customers to purchase and use Snowflake directly from Oracle.

Oracle also announced plans to introduce Nvidia's Vera Rubin NVL72 systems and HGX Rubin NVL8 systems, although the availability of these products was not specified. Oracle's record earnings in the quarter ending August 31 show that its AI investments are paying off, but concerns about an AI bubble persist.