Red Hat and AMD are expanding their collaboration to address the increasing demands of AI workloads in hybrid cloud environments. By combining Red Hat’s open-source software with AMD’s cutting-edge processors and GPUs, the two companies are offering organizations more efficient and scalable solutions for managing data, applications, and workloads across both on-premise systems and the cloud.
Powering AI Workloads with Advanced Technology
With the rise of AI, the need for data centers that can support AI workloads has never been more critical. Many existing data centers are still optimized for traditional IT operations, which leaves little room for AI-centric tasks. To address this gap, Red Hat and AMD are leveraging AMD’s Instinct GPUs alongside Red Hat OpenShift AI, offering organizations the compute power they need to run AI models efficiently.
The partnership has already proven successful in tests, such as running AMD Instinct MI300X GPUs on Red Hat Enterprise Linux AI within Microsoft Azure’s ND MI300X v5 virtual machines. These tests demonstrated that the GPUs could efficiently handle both small and large language models without the need for complex workload splitting, thus reducing both costs and operational complexity.
Streamlining AI Inference and Multi-GPU Workloads
As part of their ongoing collaboration, Red Hat and AMD are actively contributing to the upstream vLLM community, focused on optimizing AI inference. By improving kernel performance and streamlining communication, the companies aim to make multi-GPU workloads faster and more efficient. This collaboration also ensures better support for running quantized and dense AI models on AMD hardware.
A significant result of this partnership is the integration of AMD Instinct GPUs with the Red Hat AI Inference Server—a version of vLLM that has been tested and optimized for enterprise use. This allows organizations to deploy open-source AI models on AMD GPUs without requiring custom configurations, ensuring smoother deployment of AI solutions on enterprise-grade infrastructure.
Maximizing Performance with AMD EPYC CPUs
In addition to supporting GPU workloads, AMD’s EPYC CPUs are also a crucial component in the partnership. These high-performance processors are designed to handle a wide range of tasks, from AI training to inference, while also improving the overall efficiency and return on investment for GPU-powered systems. By combining the power of AMD EPYC CPUs with AMD Instinct GPUs, businesses can ensure that their infrastructure can handle diverse workloads with ease.
Simplifying Virtualization for Hybrid Cloud Adoption
Beyond AI, Red Hat and AMD are making strides in modernizing virtualization. Red Hat OpenShift Virtualization, which runs on AMD EPYC processors, enables businesses to move virtual machine-based (VM) applications to a more cloud-native platform. This solution simplifies the management of both VMs and containers within a single environment, regardless of whether they are deployed on-premise or in the public cloud.
By streamlining the management of virtualized workloads, OpenShift Virtualization can improve infrastructure efficiency and reduce overhead costs. Additionally, it supports deployment on popular server brands like Dell, HPE, and Lenovo, making it easier for businesses to integrate legacy infrastructure with modern cloud-native applications, including AI.
Building for the Future
The collaboration between Red Hat and AMD is focused on future-proofing business infrastructure by ensuring that organizations can run both AI and traditional workloads on the same systems. This flexibility allows businesses to make better use of their existing investments while preparing for the increasing demands of emerging technologies.
Ashesh Badani, Senior Vice President at Red Hat, emphasized, “To fully realize the potential of AI, organizations need the flexibility to optimize their IT footprint for growing demands. Our extended partnership with AMD offers a broader range of solutions to help businesses modernize their infrastructure, scale AI workloads, and prepare for the next generation of hardware accelerators and open-source AI technologies.”
Philip Guido, Executive Vice President at AMD, added, “As workloads become more diverse and demanding, customers require solutions that can scale. By combining Red Hat’s open-source platforms with AMD’s powerful GPUs and CPUs, we’re providing the performance and efficiency businesses need to drive innovation in AI, virtualization, and hybrid cloud environments.”
Conclusion
This expanded partnership between Red Hat and AMD represents a powerful step forward in supporting AI and hybrid cloud initiatives. By combining Red Hat’s open-source solutions with AMD’s high-performance hardware, businesses can accelerate their AI adoption, streamline virtualization, and ensure their infrastructure is ready for the future of digital transformation.
