TechForge

October 28, 2026

  • Oracle adds Google and Cerebras AI chips alongside Nvidia.
  • Oracle’s cloud revenue hits $7.4 billion as spending reaches $28.5 billion.

 

Oracle is expanding its AI chip offerings through new agreements with Google and Cerebras Systems while also expanding its partnership with Nvidia. The move will give Oracle Cloud Infrastructure (OCI) customers access to different types of AI hardware for training models and running AI applications.

The company announced the expansion during Oracle AI World. Its plans include deploying Nvidia’s Vera Rubin systems at multi-gigawatt scale, introducing Google’s TPUs, and offering specialised AI inference technology from Cerebras. Oracle is also collaborating with Intel-backed SambaNova Systems on plans to deploy SambaRack SN50 inference chips in OCI data centres, giving AI model developers another computing option.

More AI chip options for different workloads

Oracle’s agreement with Google will make OCI the first hyperscale cloud platform outside Google Cloud to offer Google’s TPUs. These processors are designed specifically for AI workloads, providing another option alongside the GPUs commonly used for training and inference.

Google’s eighth-generation TPU portfolio includes two architectures developed for different tasks. The TPU 8t is designed for training large AI models, while the TPU 8i focuses on inference, including applications involving multiple AI agents. Training involves developing AI models using large datasets, while inference refers to running trained models to generate responses.

Oracle said Google’s TPUs will support model training, fine-tuning, and inference on OCI. The company will provide the data centre infrastructure, power, cooling, and networking required to operate the processors, which are expected to become available in 2027.

“Different AI workloads have different compute needs,” Mahesh Thiagarajan, president of Oracle Cloud Infrastructure, said. He explained that providing access to multiple processor architectures would give customers more flexibility to select infrastructure based on performance, throughput, and cost requirements.

Alongside Google’s TPUs, Oracle plans to introduce Cerebras’ specialised inference technology on OCI. The partnership targets AI developers, laboratories, and enterprises building applications that require fast response times.

Inference speed is particularly relevant to AI agents that perform multiple tasks. These systems may retrieve information, execute software tools, and generate several responses before completing a request. Each interaction with an AI model can introduce delays, increasing the time needed to finish a task.

Cerebras’ Wafer-Scale Engine combines computing and memory on a single wafer-scale processor to reduce data movement between chips. According to Cerebras, the architecture supports faster token generation for AI applications involving multiple processing steps. The service is scheduled to become available on OCI in 2027.

Beyond providing additional processors, the Google partnership could also allow businesses to run AI workloads closer to data already stored in Oracle’s cloud environment. Organisations using Oracle databases could run model fine-tuning and inference on Google’s TPUs without transferring their data to another cloud provider. According to the companies, this arrangement could simplify AI deployments that depend on existing Oracle data services.

Expanding AI infrastructure with Nvidia

Oracle is also expanding its existing Nvidia-based infrastructure as it introduces processors from other suppliers. The companies announced plans to deploy Nvidia’s Vera Rubin AI systems across OCI Superclusters at multi-gigawatt scale to support demand for AI training and inference.

The companies did not disclose the financial value of the expansion. Oracle also announced that Nvidia Vera Rubin NVL72 systems and Vera CPUs are generally available on OCI. It plans to introduce Nvidia’s HGX Rubin NVL8 configuration, which uses eight Rubin GPUs and provides a smaller deployment option than the 72-GPU NVL72 systems.

The introduction of Vera CPUs expands Oracle’s Nvidia partnership beyond GPUs. While GPUs handle AI model processing, CPUs support tasks such as executing code, retrieving information, and running software tools. Oracle said customers can scale CPU and GPU resources independently, allowing them to allocate computing capacity according to their workloads.

Networking is another part of Oracle’s infrastructure expansion. The company is developing its Acceleron technology to manage data traffic and connectivity across OCI, including large AI clusters that require high-speed communication between processors and storage systems.

Oracle said its latest Acceleron developments focus on connectivity, network reliability, and workload isolation. OCI Superclusters combine Acceleron with Nvidia’s Spectrum-X Ethernet networking, Spectrum-6 switches, and BlueField-4 data processing units. Oracle also plans to use Acceleron to connect Google’s TPU clusters with OCI databases and storage services.

The financial demands of AI expansion

Oracle’s growing AI infrastructure business is also requiring substantial investment in data centre capacity and computing equipment.

In its first quarter of fiscal 2027, ended August 31, 2026, Oracle reported cloud infrastructure revenue of $7.4 billion, up 121% year over year. Total revenue increased 30% to $19.3 billion, while remaining performance obligations, which represent contracted revenue yet to be recognised, reached $664 billion.

The company also reported delivering 850 megawatts of additional data centre capacity during the quarter, reflecting the scale of its ongoing infrastructure expansion.

The investment is reflected in Oracle’s capital expenditure, which increased to approximately $28.5 billion during the quarter from $8.5 billion a year earlier. Despite the increase, the company maintained its fiscal 2027 capital expenditure forecast of $90 billion to $95 billion during its September earnings call.

Oracle has also arranged for some large AI customers to prepay for GPUs or supply the processors themselves. In June, the company said the prepaid and customer-supplied hardware portions of its large AI contracts totalled $75 billion, reducing the amount of capital it needed to raise for data centre expansion.

While Nvidia’s Vera Rubin NVL72 systems and Vera CPUs are already generally available on OCI, Google’s TPUs and Cerebras inference services are scheduled to become available in 2027.

 

 

 

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

Tech Wire Asia is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

About the Author

Muhammad Zulhusni

As a tech journalist, Zul focuses on topics including cloud computing, cybersecurity, and disruptive technology in the enterprise industry. He has expertise in moderating webinars and presenting content on video, in addition to having a background in networking technology.

Related

November 18, 2026

November 17, 2026

November 16, 2026

November 13, 2026

Join our Community

Subscribe now to get all our premium content and latest tech news delivered straight to your inbox

Popular

35684 view(s)
28211 view(s)
25638 view(s)
20740 view(s)

Subscribe

All our premium content and latest tech news delivered straight to your inbox

This field is for validation purposes and should be left unchanged.