MADRID, June 13 (Portaltic/EP) –
OpenAI, Microsoft and Oracle they have partner to use Oracle Cloud Infrastructure (OCI) for extend the Microsoft Azure AI platformin order to provide additional capacity to power OpenAI’s Artificial Intelligence (AI) services.
The company specialized in AI led by Sam Altman offers Generative AI services to more than 100 million users every month, This implies a practically unlimited demand for AI infrastructures to carry out the necessary workloads.
Currently, the American technology company makes use of Microsoft infrastructures to drive these workloads, however, in order to Gain access to more infrastructure and continue advancing AI capabilities, OpenAI and Microsoft have partnered with Oracle, with the idea of extending the Microsoft Azure AI platform to the Oracle OCI infrastructure and thus providing additional capacity for OpenAI.
As Oracle’s president and chief technology officer, Larry Ellison, explained in a statement., faced with a panorama in which “the race to build the world’s largest language model is underway,” “a” is being created unlimited demand” for AI infrastructureas is the case of infrastructure Oracle Gen 2.
Thus, OpenAI will make use of this Oracle Gen 2 OCI to be able to extend its AI capabilities in an infrastructure that, as defined by Ellison “is the fastest and most profitable in the world. Specifically, it is an infrastructure that has AI capabilities designed specifically for OICand allow companies create and train models “faster and more reliable” anywhere in the Oracle cloud.
Following this line, the company has detailed that to train large language models (LLM), the OIC Supercluster allows you to scale up to NVIDIA Blackwell GPU of 64k either Grace Blackwell GB200 Superchips connected through a network of RDMA cluster ultra-low latency and an option to HPC storage.
Additionally, you have OCI Compute virtual machines and OCI bare metal Nvidia GPU instances, which can driving applications for generative AI, computer vision, natural language processing, and recommendation systemsamong other capabilities.
With all this, the CEO of OpenAI, Sam Altman, has stressed that, thanks to the partnership with Microsoft and Oracle, as well as the capabilities offered by the OCI, OpenAI will be able to “continue to scale.”
Likewise, with this association, OpenAI will join “thousands of AI innovators in industries around the world,” who already run their AI workloads on OCI infrastructure. Among these companies are some such as Modal, Adept, Nvidia, Reka, Twelve Labs and xAI by Elon Musk.
However, OpenAI has clarified in a post on X (former Twitter) that its Strategic cloud relationship with Microsoft “has not changed” and that, with the partnership with Oracle, the company will use the Azure AI platform for “inference and other needs.” Therefore, all the Pre-training of OpenAI models “will continue to be performed on supercomputers built in association with “Microsoft.”
