Ever since the arrival of generative artificial intelligence (AI) platforms like ChatGPT and Bard a year ago, the race among big tech companies to embrace this transformative tech has only intensified. Global software giant Oracle Corporation, too, is making strides by tying up with companies like NVIDIA and Elon Musk’s xAI.
Oracle, like many of its peers, has been using AI for quite some time, but “this [generative AI] is different,” says Founder Larry Ellison. “It is a revolution, a breakthrough; it’s transformational. It’s fundamentally changing things at Oracle,” he says. Ellison, who has closely seen technology evolve through the years, is one of the few people who can say this with authority. He co-founded Oracle in 1977; and even at the age of 79, serves as its Chairman and CTO.
Oracle has been competing with the likes of Amazon Web Services, Google Cloud, Microsoft Azure, and others in the cloud space. However, some big names developing AI have recently signed contracts with the Texas-based tech giant to purchase more than $4 billion worth of capacity on Oracle Cloud Infrastructure (OCI) to train their generative AI models. This is double the amount Oracle earned in the fourth quarter ended May 31, 2023.
The company attributes this growth to “better speeds” and the “price advantage” it offers to clients. For example, Oracle uses a remote data memory access (RDMA) network for its Generation 2 Cloud, which allows one computer to access the memory of another without tapping on its shoulder and interrupting it. This allows for data transfer between computers that is exponentially faster than conventional networks. Further, its new Oracle Cloud Infrastructure Compute instances powered by NVIDIA H100 Tensor Core and NVIDIA L40S graphics processing units (GPUs), combined with Ampere AmpereOne CPUs also help it deliver improved speed and pricing.
Explains Ellison: “When we interconnect those GPUs, that computer runs much faster in our cloud than it does in other clouds. And in the cloud, time is money. If we run twice as fast, it costs half as much. If we run three times faster, we cost a third as much.”
For this reason, many big names, including NVIDIA, Cohere and xAI, are leveraging Oracle Cloud to train their generative AI models. Ellison is convinced that one of the NVIDIA superclusters the company is building will be the world’s largest scientific computer. It’s no surprise then that Oracle’s AI strategy is becoming a primary factor in its future expansion plans.
@nidhisingal
The writer was in Las Vegas at the invitation of Oracle Corp.