AMD and Rackspace Technology have entered into a new partnership focused on enterprise AI infrastructure. The agreement outlines a phased deployment of a 30-megawatt footprint dedicated to AMD-based compute solutions within Rackspace's global data centres, scheduled to begin between late 2026 and 2028. The partnership builds on a previously established Memorandum of Understanding and expands the existing relationship between the two companies within Rackspace’s AI ecosystem.
The proposed deployment is intended to support enterprise workloads, including those in sectors such as healthcare that have expressed interest in accelerated computing capabilities. Using AMD's EPYC CPUs and Instinct GPUs, Rackspace plans to incorporate these technologies into its Enterprise AI Cloud architecture. The architecture is designed to manage workloads by allocating tasks to appropriate computing resources while providing performance monitoring and operational oversight throughout execution.
Under the collaboration, AMD and Rackspace plan to offer managed AI infrastructure through a single provider rather than relying on multiple vendors for different components. The approach is intended to address the requirements of organisations in regulated industries that need oversight, compliance controls, and operational efficiency from AI platforms. The planned infrastructure will cover services ranging from bare metal computing to managed inference engines, providing organisations with an integrated AI platform.
Both companies also intend to work together to identify and support enterprise customers. Sales and marketing resources from AMD and Rackspace are expected to be used to promote the adoption of AMD-powered infrastructure, particularly within regulated sectors. The partnership is also expected to support the deployment of agentic workflows within enterprise systems.
The stated objective of the partnership is to provide governed and accountable infrastructure that can support the transition of AI initiatives from experimental projects to operational business workloads. As enterprise adoption of AI continues to expand, partnerships such as this are intended to address growing infrastructure and computing requirements.