Dive Brief:
- Oracle is racing to expand cloud capacity in anticipation of a demand surge, the company’s CEO said Monday during a quarterly earnings call. “Because we have far more demand than we can supply, our biggest challenge is building data centers as quickly as possible,” Safra Catz said.
- The company has 64 active cloud regions globally, 44 of which support public cloud. Six more are currently under construction, Catz said.
- After outpacing larger cloud providers in the first half of the year, Oracle’s cloud revenue growth slowed during the three months ending Aug. 31, according to the company’s Q1 2024 earnings report. Oracle saw combined IaaS and SaaS revenues increase by 30% year over year for the quarter, down from a 54% bump in the previous quarter.
Dive Insight:
As an underdog hyperscaler, Oracle has banked on building more data centers in more places than its larger competitors. The strategy promises to deliver lower latency and greater data control to meet data sovereignty requirements.
“We believe we just have to get into more countries than Amazon, let’s say, because we have to serve those countries where we have a large installed base,” Oracle chairman and CTO Larry Ellison said in a 2020 blog post.
In June, the company opened its first two European cloud regions to satisfy EU privacy and security regulations. The Frankfurt, Germany, and Madrid data centers maintain physical separation from other Oracle cloud regions, ensuring workloads remain in their jurisdiction of origin.
The build out has been accompanied by systems upgrades, Ellison said Monday.
“We're on our second generation of data center and our second generation of cloud,” Ellison said. “We moved from a generation which we were not very happy with to a second generation, which we think solved a lot of problems.”
Like AWS, Microsoft and Google Cloud, Oracle is optimizing its cloud infrastructure to handle a generative AI-fueled surge in enterprise usage. In October, Oracle partnered with chip manufacturer Nvidia to beef up its AI-processing capacity.
AI development companies have already commissioned more than $4 billion in cloud-based AI training capacity from Oracle, according to Ellison. “That's twice as much AI training as we had booked at the end of the last quarter,” Ellison said, noting that Elon Musk’s xAI is among the growing customer list.
In addition, Nvidia extended the Oracle partnership to include the chip designer’s cloud-based AI supercomputing service. The service leverages Oracle’s Supercluster infrastructure, built on Nvidia GPUs.
“Is generative AI the most important new computer technology ever?,” Ellison asked during the earnings call. “Maybe. We're about to find out.”