Oracle Cloud Infrastructure customers are set to receive an expanded selection of compute options.
A significant development is the integration of Google's Tensor Processing Units (TPUs) into OCI. This arrangement positions Oracle as the first hyperscale cloud platform outside of Google's own services to offer these specific chips. Availability of these TPUs within OCI is slated to commence in 2027, with capacity scaling across subsequent years.
This move suggests Oracle is actively seeking to reduce its dependence on Nvidia, with whom the company has secured multi-billion dollar agreements for its data center buildouts, including the planned Texas facility with the GPU manufacturer and OpenAI.
OCI president Mahesh Thiagarajan stated, "Different AI workloads have different compute needs." He added, "Bringing Google TPUs to OCI gives customers … the flexibility to choose the right infrastructure for the performance, throughput, and economics of each AI workload."
Simultaneously, Oracle is advancing its collaboration with OpenAI, introducing new ChatGPT integrations on Tuesday. Two new offerings incorporate models from the House of Altman. Oracle AI Workspace functions as a "single user experience that brings OpenAI frontier models together with Oracle applications, enterprise data, and cloud services" via ChatGPT plugins, effectively making Oracle's existing platforms accessible through ChatGPT.
For enhanced security, Oracle AI Security Mesh is another introduction. This managed platform merges frontier AI models for cybersecurity with Oracle's established security proficiency. It will utilize OpenAI cybersecurity models to identify weaknesses, test exploitability, and suggest remedies across authorized components of a corporation's technology stack.
Beyond AI integration, Oracle announced a native service partnership with Snowflake, allowing customers to procure and utilize the AI data cloud directly from Oracle, managing all associated billing and discounts. Furthermore, Oracle Universal Credits are applicable to this new Snowflake offering on OCI.
Although diversifying, Oracle also unveiled new Nvidia hardware in Las Vegas. The company announced the general availability of Nvidia Vera Rubin NVL72 systems and Nvidia Vera CPUs, alongside plans to introduce HGX Rubin NVL8 systems, though an exact timeline for their release was not provided. Thiagarajan remarked, "By combining NVIDIA Vera Rubin GPUs and Vera CPUs with Oracle Acceleron networking and AI-optimized data centers, we’re helping customers keep GPUs productive, secure, and scale efficiently as their needs grow."
These infrastructure updates occur as Oracle reported record earnings for the quarter ending August 31, with revenue increasing 30 percent year over year, indicating that prior AI investments are beginning to yield returns. This rapid growth in certain firms prompts scrutiny regarding the sustainability of current AI market valuations.
