WORLDTECHOracle to offer Google TPUs in its cloud from 2027
Oracle plans to make Google's Tensor Processing Units available through Oracle Cloud Infrastructure, according to a report by Crypto Briefing. The chips are set to debut on OCI in 2027, and the arrangement would make Oracle the first major cloud provider outside Google to offer them.
Announcement and timing
The partnership was unveiled on October 27, 2026, at Oracle AI World in Las Vegas. The two companies said Google's Tensor Processing Units, or TPUs, will reach OCI in 2027, with a phased rollout expected to build over several years.
Which chips are involved
The hardware comes from Google's latest generation, which includes one model built for training and another built for inference. Training is the costly stage in which a model learns from data, while inference is the stage in which a finished model responds to requests at scale. One of the chips is described as delivering nearly three times the compute performance per pod of the previous generation, and a pod is a large cluster of chips linked to operate as a single system.
Oracle's broader chip strategy
Oracle's AI chip lineup has relied heavily on Nvidia GPUs, so TPUs give the company a second major option to put in front of customers. Oracle is also deepening its work with Nvidia, aiming for multi-gigawatt scale with Nvidia's next-generation Vera Rubin systems, and is bringing in hardware from Cerebras Systems and SambaNova Systems.
Market reaction and wider significance
Offering the chips makes OCI the first major cloud provider outside Google to carry them, widening the choice of AI hardware and possibly shifting competition among cloud providers. According to the reporting, investors responded positively to the announcement.
Open questions
- How quickly will the phased rollout of TPUs on OCI progress after 2027?
- What share of Oracle's AI workloads will move to TPUs compared with Nvidia hardware?