Oracle will offer Google’s Tensor Processing Units (TPUs) through Oracle Cloud Infrastructure (OCI) starting in 2027. This will be the first time Google’s TPUs are made available through a hyperscale cloud provider outside Google.
According to a press release, OCI will provide the data center infrastructure, including power, cooling and networking, required to operate the processors and deployment will take place over several years under the agreement. The TPUs will be used for AI model training and inference, including workloads involving large language models, mixture-of-experts (MoE) models and applications that use multiple AI agents.
Mahesh Thiagarajan, president, Oracle Cloud Infrastructure, Oracle, said, “Different AI workloads have different compute needs, bringing Google TPUs to OCI gives customers access to multiple leading AI compute architectures in one cloud. That choice gives them the flexibility to choose the right infrastructure for the performance, throughput, and economics of each AI workload.”
Amin Vahdat, senior vice president and chief technologist, AI & Infrastructure, Google, said, “The next generation of AI will require a diverse range of compute, TPUs give customers a powerful, purpose-built option for demanding training and inference workloads. Bringing Google TPUs to OCI will give more organizations access to this technology and greater flexibility in how they build and run their most complex AI applications.”
Google’s eighth-generation TPU range includes two models: TPU 8t, designed for model training, and TPU 8i, designed for inference. The processors are scheduled to become available through OCI in 2027 although the companies have not specified the initial deployment capacity or the full rollout schedule.