The surge of agreements surrounding advanced artificial intelligence processors shows no sign of slowing, and this latest development centers on a newly forged alliance between AMD and Oracle. On Tuesday, the two companies revealed their intention for AMD to integrate an immense deployment—approximately 50,000 units—of its most sophisticated AI chipsets into Oracle-operated data centers. This large-scale rollout is expected to begin during the latter half of 2026, marking a major milestone in AMD’s continuing expansion into high-performance computing infrastructure.
Although the announcement confirmed the technical and logistical aspects of the arrangement, both companies refrained from disclosing any financial details associated with the partnership. The lack of immediate comment to inquiries from Business Insider further reinforced the confidential nature of the deal’s economic terms, a common occurrence in agreements of this magnitude within the technology sector.
This collaboration emerges a mere eight days after AMD revealed another headline-grabbing agreement—a multibillion-dollar partnership with OpenAI for the supply of its chips. The juxtaposition of these two announcements highlights AMD’s accelerating momentum in the increasingly competitive AI hardware landscape. Collectively, the deals significantly fortify AMD’s strategic position against its principal rival, Nvidia, a company widely recognized for its longstanding dominance in the advanced GPU market.
In this new arrangement, Oracle is set to integrate AMD’s Instinct MI450 graphics processing units—GPUs renowned for their superior computational throughput and energy efficiency. Notably, these are the same chip models scheduled for deployment by OpenAI in its AI training and inference operations. Such GPUs are critical components in the training of large-scale machine learning models, as well as in the inference process—the stage at which those models utilize their acquired knowledge to interpret data and perform complex analytical tasks.
During an interview with CNBC on Tuesday, Karan Batta, senior vice president of Oracle Cloud Infrastructure, expressed considerable optimism regarding AMD’s potential impact in this field. He emphasized that client demand for AMD’s technology is anticipated to be particularly strong in the inference domain, where efficiency, latency reduction, and scalability are all crucial performance metrics.
In the hours following the announcement, AMD’s shares reflected a modest yet positive reaction, rising roughly two percent in premarket trading. The increase, however, was comparatively subdued when measured against the dramatic twenty-seven percent surge experienced after the company’s earlier agreement with OpenAI. Oracle’s shares, conversely, experienced a slight decline of about one and a half percent, likely reflecting broader market adjustments rather than a negative response to the news itself.
This partnership between Oracle and AMD joins a growing list of ambitious AI infrastructure initiatives unveiled by major technology corporations in recent months. On Monday, Broadcom and OpenAI formally confirmed a monumental plan to deliver ten gigawatts of custom-designed AI accelerators—specialized processors optimized for machine learning workloads. This follows similar announcements by Nvidia last month, in which the chipmaking titan disclosed that OpenAI would be granted access to an equivalent ten gigawatts of its GPUs, backed by a staggering one-hundred-billion-dollar investment.
Adding yet more context to this rapidly evolving competitive environment, back in July, OpenAI and Oracle announced a far-reaching collaboration aimed at constructing up to 4.5 gigawatts of AI computational capacity—an undertaking valued at more than three hundred billion dollars over five years. Viewed in aggregate, these initiatives demonstrate the scale and urgency with which companies across the technology spectrum are racing to build the infrastructure needed to sustain the next generation of artificial intelligence innovation.
Sourse: https://www.businessinsider.com/amd-oracle-deal-ai-chip-wars-gpus-nvidia-openai-2025-10
