Subscribe

Cloud

Oracle to Offer Google TPUs on OCI in 2027 and Build ChatGPT and Codex Into Its Apps

At Oracle AI World, Oracle said it will be the first hyperscaler outside Google to offer Google's AI chips, and expanded its OpenAI partnership across its applications and cloud.

DailyMech Staff

Published

Last updated / Fact-checked on

Oracle to Offer Google TPUs on OCI in 2027 and Build ChatGPT and Codex Into Its Apps
Photo: Håkan Dahlström / Wikimedia Commons — CC BY 2.0

Oracle and Google announced on Oct. 27 at Oracle AI World in Las Vegas that Google's Tensor Processing Units will become available through Oracle Cloud Infrastructure (OCI). It is the first time Google TPUs will be offered on a hyperscale cloud platform outside Google itself. OCI will supply the power, cooling and data center networking needed to run the chips at scale. Availability is planned for calendar year 2027, with capacity ramping over several years.

Customers will be able to choose TPUs within OCI for large-model training and post-training, including dense and mixture-of-experts models, and for production inference in reasoning, agentic and long-context applications. Google's eighth-generation lineup has two designs: TPU 8t for training and TPU 8i for inference. The companies say TPU 8t cuts frontier model development from months to weeks and delivers nearly three times the compute performance per pod of the previous generation, while TPU 8i is built for the iterative work of many specialized agents.

"Different AI workloads have different compute needs," said Mahesh Thiagarajan, president of Oracle Cloud Infrastructure. "Bringing Google TPUs to OCI gives customers access to multiple leading AI compute architectures in one cloud." Amin Vahdat, Google's senior vice president and chief technologist for AI and infrastructure, said the next generation of AI will require a diverse range of compute.

OpenAI comes to Oracle's apps

The same day, Oracle announced an expanded collaboration with OpenAI that brings OpenAI models, ChatGPT and Codex together with Oracle's applications, data and cloud. It includes four offerings.

Oracle AI Workspace combines OpenAI frontier models with Oracle applications, enterprise data and cloud services. Through plugins in ChatGPT, employees can look up information, complete approved workflows and take action in Oracle systems, including Fusion Applications, NetSuite, Oracle AI Database and OCI, while keeping enterprise governance and approval controls.

Oracle AI Security Mesh is a managed service that uses OpenAI cyber models to help customers find and fix vulnerabilities across OCI, other clouds and on-premises environments. Oracle runs the testing infrastructure and safeguards; customers define what may be tested.

Oracle App Composer lets employees turn ideas into enterprise applications without coding expertise. Users build with Codex while Oracle handles deployment and operations on OCI, with enterprise identity, permissions and audit controls applied.

Oracle AI SMB Bundle is a free-to-start package for small businesses that includes $12,000 in OCI credits per year for up to three years, plus ChatGPT Business, Oracle AI Workspace, Codex, App Composer and NetSuite QuickStart for invoicing and cash flow. Eligibility rules apply; applicants must be approved, privately held U.S. businesses.

"OpenAI is transforming how enterprises solve business problems, from chat, to coding, to analytics, to autonomous agents," said Clay Magouyrk, CEO of Oracle. Greg Brockman, OpenAI's president and co-founder, said the partnership helps bring frontier AI into everyday operations.

Why it matters

The announcements come as enterprises look for more choice in AI compute and tighter integration of AI into the systems they already run. Oracle is widening its chip lineup beyond Nvidia: Reuters reported that it also plans to offer Cerebras systems, while it expands its Nvidia Vera Rubin deployments to multi-gigawatt scale. Offering Google's TPUs on a rival cloud is a notable sign that Nvidia's grip on AI compute is loosening, though the TPU capacity won't arrive until 2027 and the OpenAI products are still rolling out.

Sources

In Cloud

From other sections

Sections

The Daily BriefingSample

Intelligence, distilled. Every morning.

Sample briefing. A five-minute read of the most important stories in AI and technology, once this list is real.

Sample form. It does not work. The address is not stored or sent.