Red Hat and AMD expand collaboration for AI and virtualization at Red Hat Summit 2025

Red Hat and AMD announced a strategic collaboration at the Red Hat Summit to enhance AI capabilities and virtualized infrastructure across hybrid cloud environments. The partnership focuses on integrating Red Hat’s open source solutions with AMD’s computing architectures to support AI workloads and modernize virtual machine (VM) deployments.
The collaboration enables AMD Instinct GPUs on Red Hat OpenShift AI for AI deployments across hybrid clouds. Testing on Microsoft Azure ND MI300X v5 using AMD Instinct MI300X GPUs with Red Hat Enterprise Linux AI demonstrated AI inferencing for small and large language models on a single VM, reducing the need for multiple VMs. Red Hat and AMD are also contributing to the vLLM community to improve AI inference on AMD GPUs by optimizing the AMD kernel library, Triton kernel, and FP8 components for dense and quantized models, enhancing multi-GPU support for scalable workloads, and expanding vLLM ecosystem engagement with partners like IBM. AMD Instinct GPUs will support Red Hat AI Inference Server, a vLLM distribution, for deploying open source AI models on validated GPU hardware. AMD EPYC CPUs are also utilized to host GPU-enabled systems for AI workloads.

Ashesh Badani, senior vice president and chief product officer, Red Hat said “Fully realizing the benefits of AI means that organizations must have the choice...