TLDR: Oracle announced a significant expansion of its partnership with AMD, planning to deploy 50,000 of AMD’s upcoming Instinct MI450 series AI chips starting in the third quarter of 2026. This strategic move aims to bolster Oracle Cloud Infrastructure’s (OCI) capabilities for large-scale AI workloads, including generative AI and large language models, while diversifying its AI chip supply chain and reducing reliance on NVIDIA.
Oracle has unveiled a major expansion of its collaboration with AMD, committing to deploy 50,000 of AMD’s cutting-edge Instinct MI450 series artificial intelligence (AI) chips within its cloud infrastructure. This initiative is set to commence in the third quarter of 2026, with plans for further expansion beyond 2027. The announcement, made on October 15, 2025, highlights Oracle’s intent to become the first hyperscaler to publicly offer infrastructure powered by these advanced AMD GPUs.
The primary objective of this partnership is to significantly scale Oracle’s AI capabilities, enabling customers to develop and deploy ambitious AI applications. The planned large-scale AI superclusters will leverage AMD’s new “Helios” rack design, incorporating Instinct MI450 GPUs, next-generation EPYC CPUs codenamed Venice, and Pensando networking technology codenamed Vulcano. The MI450 chips are notable as AMD’s first AI processors designed to be integrated into larger rack-sized systems, allowing up to 72 chips to function as a single, cohesive unit. Each AMD Instinct MI450 GPU is expected to offer up to 432 GB of HBM4 and 20 TB/s of memory bandwidth, facilitating the execution of models 50 percent larger than previous generations entirely within memory.
This collaboration is a strategic step for Oracle to diversify its AI chip supply chain, lessening its dependence on NVIDIA, which currently dominates the AI chip market with over 90% market share. Mahesh Thiagarajan, Executive Vice President of Oracle Cloud Infrastructure, emphasized the importance of robust, scalable, and high-performance infrastructure for customers building advanced AI applications. He stated, “By bringing together the latest AMD processor innovations with OCI’s secure, flexible platform and advanced networking powered by Oracle Acceleron, customers can push the boundaries with confidence.”
Forrest Norrod, Executive Vice President and General Manager of AMD’s Data Center Solutions Business Group, underscored the collaboration’s focus on advancing customer AI initiatives. He noted, “With our AMD Instinct GPUs, EPYC CPUs, and advanced AMD Pensando networking, Oracle customers gain powerful new capabilities for training, fine-tuning, and deploying the next generation of AI.” He added that together, AMD and Oracle are “accelerating AI with open, optimized, and secure systems built for massive data centers.”
The partnership builds on a long-standing collaboration between Oracle and AMD, which previously saw OCI adopt AMD Instinct MI300X GPU-powered shapes in 2024 and the release of AMD Instinct MI355X GPUs for OCI compute earlier this year. The expanded infrastructure is specifically intended to support customers engaged in training large language models, running generative AI workloads, and advancing high-performance computing use cases.
Also Read:
- Oracle Unveils Global AI Data Platform, Bolstered by $1.5 Billion in Partner Investments
- Cerebras Systems to Power UAE’s Stargate AI Supercomputing Hub Amid Global Expansion
This deal also reflects a broader trend in the tech industry, where major players are increasingly turning to AMD as an alternative to NVIDIA for AI development. For instance, OpenAI recently announced a separate agreement with AMD for AI chips, requiring a total of 6 gigawatts of power until 2029. The strengthening cooperation between big tech and AMD is also anticipated to positively impact suppliers like Samsung Electronics, which currently provides high-bandwidth memory (HBM) to AMD for its MI350 AI chip, with a high possibility of supplying HBM for the MI450 series as well.
