Oracle is expanding its artificial intelligence (AI) infrastructure partnerships with Nvidia while diversifying its cloud computing offerings to include chips from competitors such as Alphabet’s Google and Cerebras Systems.

Oracle and Nvidia announced plans to scale their collaboration to “multi-gigawatt scale” using Nvidia’s latest Vera Rubin AI server systems. The companies did not disclose the financial details of the agreement. However, Nvidia CEO Jensen Huang has previously indicated that Vera Rubin systems could generate up to $40 billion in revenue per gigawatt for the chipmaker.

Despite strengthening its relationship with Nvidia, Oracle is pursuing a strategy that gives customers access to a wider range of AI computing technologies. The company said it would use its proprietary Acceleron networking chips in certain Nvidia-powered systems instead of some of Nvidia’s networking products.

On Tuesday, Oracle also announced that it would begin offering Google’s tensor processing units (TPUs) through its cloud infrastructure for the first time. The move reflects growing demand for specialised chips designed to handle different AI workloads efficiently.

“Various AI tasks require distinct computing requirements,” said Mahesh Thiagarajan, president of Oracle Cloud Infrastructure. He added that the approach would allow customers to choose infrastructure based on performance, throughput and cost-effectiveness.

Oracle further plans to offer chips from SambaNova Systems and Cerebras Systems, both supported by Intel. These companies develop processors designed for AI inference, the stage in which trained AI models respond to user requests.

Also Read: From Silos to Synergy: Building the Future With Agentic AI

While Nvidia remains a dominant player in AI training hardware, the inference market is attracting increasing competition as businesses seek faster and more cost-effective ways to run AI applications.

Cerebras CEO Andrew Feldman told Reuters that Oracle’s move represents a significant expansion of the company’s customer base among major clients. He said the partnership demonstrates that Cerebras technology is being deployed at scale in other companies’ data centres.

Also Read: India's Deep-Tech Revolution: 10 AI Startups Driving Innovation

Oracle’s broader strategy highlights the growing importance of chip diversity as cloud providers compete to deliver flexible, efficient and affordable AI computing services.