Oracle to Launch AI Supercluster with 50,000 AMD GPUs in 2026

Oct 21, 2025
Oracle and AMD announced an expanded partnership that will see Oracle Cloud Infrastructure deploy 50,000 AMD Instinct MI450 GPUs starting in Q3 2026. The new AI supercluster will use AMD’s Helios rack design and next-generation EPYC CPUs.

Oracle Corporation and Advanced Micro Devices, Inc. (AMD) announced an expansion of their collaboration to deliver large-scale AI infrastructure, according to a press release. Starting in the third quarter of 2026, Oracle Cloud Infrastructure (OCI) will deploy 50,000 AMD Instinct MI450 GPUs, forming what it says will be the first publicly available AI supercluster powered by AMD’s latest hardware.

The deployment will use AMD’s “Helios” rack design, combining the Instinct MI450 GPUs with next-generation EPYC CPUs codenamed “Venice” and Pensando networking technology known as “Vulcano.” The architecture is designed to support large-scale AI training and inference tasks with improved performance and energy efficiency.

Each MI450 GPU offers up to 432 GB of HBM4 memory and 20 TB/s of bandwidth, enabling larger in-memory AI models compared to prior generations. The system also includes DPU-accelerated networking and support for the open-source ROCm software stack, which provides compatibility with widely used AI frameworks.

The companies plan to expand the cluster further in 2027 and beyond, adding to Oracle’s existing AMD-based infrastructure, including previously deployed MI300X and MI355X GPU clusters. The new system will be part of OCI’s zettascale supercluster lineup, aimed at meeting growing demand for high-capacity AI computing.

We hope you enjoyed this article

Consider subscribing to one of our newsletters like Silicon Brief or Daily AI Brief.

Also, consider following us on social media:

AI Chips & Datacenters AI Brief AI Brief (X)
Free newsletter

Silicon Brief

Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.

AI Token Webinar

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more

You may also like

Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.