Oracle moves into AI infrastructure

Market coverage says Oracle is positioning OCI Gen 2 and AI 'Superclusters' as a go‑to provider for hyperscale AI workloads and to support Nvidia GPU deployments, suggesting hiring and projects may cluster around Oracle's infrastructure. Coverage also emphasizes Nvidia's central role in data‑centre revenue amid rising cloud demand for GPU capacity. (markets.financialcontent.com, markets.financialcontent.com)

Oracle is no longer pitching itself mainly as a database company. By April 2026, it is selling Oracle Cloud Infrastructure as a place to train and run giant artificial intelligence systems built on Nvidia chips. (oracle.com) Oracle’s latest numbers show why that pitch is getting attention. In the quarter ended February 28, 2026, Oracle said cloud infrastructure revenue rose 84% to $4.9 billion, while total cloud revenue reached $8.9 billion and remaining performance obligations climbed 325% to $553 billion. (oracle.com) The company has spent the past year tying that growth to bigger computing clusters. Oracle said in October 2025 that its Oracle Cloud Infrastructure Zettascale10 system would connect hundreds of thousands of Nvidia graphics processing units across multiple data centers and deliver up to 16 zettaFLOPS of peak performance. (oracle.com) Those systems are built for a simple problem: training large artificial intelligence models takes vast numbers of graphics chips linked together with very fast networking. Oracle and Nvidia said on March 17, 2026 that they were expanding their partnership to add new Oracle Cloud Infrastructure supercomputing and vector-search capabilities aimed at moving customers from experiments into production. (blogs.oracle.com) Oracle is also tying that infrastructure push to named projects. OpenAI said on September 23, 2025 that new Stargate sites being developed by Oracle would join the Abilene, Texas campus already running on Oracle Cloud Infrastructure, bringing Stargate to nearly 7 gigawatts of planned capacity and more than $400 billion in investment over three years. (openai.com) That helps explain why Nvidia sits at the center of the story. Nvidia reported on February 25, 2026 that quarterly data center revenue hit $62.3 billion, up 75% from a year earlier, out of total quarterly revenue of $68.1 billion. (nvidianews.nvidia.com) Oracle is not the only cloud provider chasing that spending. Amazon Web Services, Microsoft Azure, Google Cloud, and specialist provider CoreWeave are all competing for the same market for rented graphics processing units, but Oracle has been arguing that its newer cloud network design and large dedicated clusters can win a share of the biggest workloads. (nvidia.com) “Gen 2” is Oracle’s label for the cloud architecture it has been building for several years, with more isolation between control systems and customer workloads than older cloud designs. The phrase is old, but Oracle is now using that platform as the base for its newer artificial intelligence superclusters rather than as a side business to databases and business software. (blogs.oracle.com, oracle.com) Investors have started marking the shift in the stock. Oracle shares rose as much as 10% in extended trading on March 10, 2026 after the company reported results and raised its outlook for fiscal 2027. (cnbc.com) The next test is whether Oracle can turn signed demand into running capacity fast enough. Its backlog, its Nvidia tie-up, and its role in Stargate all point in the same direction: more of Oracle’s future now depends on building data centers, not just selling software. (oracle.com, openai.com)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.