Oracle‘s latest earnings call revealed a cloud strategy that’s less about chasing Nvidia GPUs and more about a quieter advantage the company has held for decades: roughly 80 percent of the world’s largest companies already keep their data inside Oracle databases, according to Gartner figures cited by longtime Oracle cloud partner Scott Morrell. That fact is doing more work in Oracle’s cloud AI pitch than any single infrastructure announcement.
The raw numbers behind Oracle Cloud Infrastructure are substantial on their own. Oracle delivered 850 megawatts of AI capacity and more than 300,000 GPUs to cloud customers since the fourth quarter, according to co-CEO Clay Magouyrk, and GPU utilization across the cloud fleet sat at 97.9 percent in the first quarter. GPUs coming up for renewal got resold or renewed at a 20 percent premium over their prior contracts, which suggests demand for Oracle’s cloud capacity isn’t slowing even as competitors race to build their own.
Morrell’s point, though, cuts closer to why enterprises actually pick a cloud provider for AI work in the first place. Moving massive datasets into whichever cloud happens to run a given AI model gets expensive fast, between egress charges and the security risk of shuttling sensitive data around. Oracle is betting companies would rather bring AI models to data that’s already sitting in their cloud than move the data instead, and Magouyrk described that strategy plainly: placing Oracle’s AI database inside every major cloud, then connecting it to enterprise data, applications and developer workflows already in place.
That approach is showing up in Oracle’s cloud application business too. Customers ran Oracle’s embedded cloud AI more than 150 million times in the quarter, and AI agents executed more than 3.5 million times in production, according to co-CEO Mike Sicilia. NetSuite’s cloud AI connector service reached more than 10,000 customers, making it one of the fastest-adopted tools in the platform’s history.
The financials back the strategy for now. Oracle posted $19.35 billion in quarterly revenue, up 30 percent year over year, with cloud revenue alone jumping 62 percent to $11.61 billion. Whether Oracle’s data-gravity edge holds as rivals build out their own cloud AI infrastructure is the real question, but for now, Oracle isn’t trying to out-GPU anyone. It’s betting the data was already sitting in the right cloud all along.
