Connect with us

Tech

AMD Secures Major AI GPU Deal with Oracle Amid Nvidia Dominance

Published

on

Amd Mi355x Ai Gpu Oracle Cluster

San Jose, CaliforniaAMD has recently made waves in the AI hardware market by securing a multi-billion dollar contract with Oracle to provide a cluster of 30,000 MI355X AI accelerators. This deal, announced during Oracle’s Q2 2025 earnings call by Chairman Larry Ellison, marks a significant step for AMD as it seeks to challenge Nvidia‘s current dominance, which holds approximately 95 percent of the GPU market.

The MI355X chips, manufactured on TSMC’s 3nm node and based on AMD’s CDNA 4 architecture, are expected to ship by mid-2025. Each accelerator will boast 288GB of HBM3E memory and a bandwidth capacity of up to 8TB per second, explicitly supporting FP6 and FP4 low-precision computing. These specifications position the MI355X as a formidable competitor to Nvidia’s line of GPUs, including the Blackwell models, according to TechRadar.

Ellison highlighted the strategic importance of this deal, stating, “In Q3, we signed a multi-billion dollar contract with AMD to build a cluster of 30,000 of their latest MI355X GPUs.” The contract strengthens Oracle’s position in the rapidly growing AI cloud market, alongside previous shipments of MI300X GPUs to both Oracle and Vultr.

The Oracle-AMD partnership also supports Project Stargate, a massive AI data center initiative that was announced earlier this year. The project is poised to involve a liquid-cooled Nvidia GB200 cluster with 64,000 GPUs aimed at AI training. Ellison elaborated, “Stargate looks to be the biggest AI training project out there, and we expect that will allow us to grow our RPO even higher in the coming quarters.”

Oracle’s innovative AI data platform is designed to enable its extensive database customer base to leverage advanced AI models from companies like OpenAI, XAI, and Meta. This transition allows clients to convert their existing data into vector formats compatible with AI models, enhancing data analysis capabilities while maintaining security. Ellison emphasized the economic advantages of Oracle’s approach, asserting, “If you run faster and pay by the hour, you cost less. So that technology advantage translates to an economic advantage which allows us to win a lot of these huge deals.”

Despite facing stiff competition from Nvidia, AMD’s recent strides in the AI chip sector demonstrate its growing significance. The successful implementation of the MI355X accelerators in Oracle’s AI projects could further reshape the landscape of cloud computing and AI training.

1x