AMD (global semiconductor company) and Rackspace Technology (cloud computing company) have signed a definitive agreement for the phased deployment of 30 megawatts of AMD-based compute infrastructure across Rackspace's global data center footprint, with rollout beginning in late 2026 and extending through 2028.

The deal, announced, converts a Memorandum of Understanding the two companies signed on May 7, 2026, into a binding commercial framework and formally establishes AMD as a strategic technology partner at the silicon layer of Rackspace's governed AI stack.

Scope and Timeline of the Deployment

The agreement covers an initial 30 MW footprint dedicated exclusively to AMD-based compute. According to the companies, at full deployment, that capacity will be directed toward regulated enterprise workloads, with healthcare providers specifically cited as having expressed early interest in accelerated compute for clinical AI and inference at scale.

The phased nature of the rollout means individual deployment authorizations remain subject to separate execution, and the companies cautioned in their forward-looking statements that the full 30 MW may not be achieved in its entirety or may proceed on a timeline that differs from the current plan.

The hardware at the center of the agreement includes AMD Instinct GPUs, specifically the MI355X and MI350P models, along with future successor solutions, as well as AMD EPYC CPUs.

Both product lines will be integrated into what the companies describe as an Enterprise AI Cloud architecture, designed to route each workload to the appropriate compute resource while maintaining what Rackspace characterizes as full accountability for performance and outcomes.

Four Integrated Capabilities

The definitive agreement is structured to accelerate delivery of four integrated capabilities that were first outlined when the MOU was announced in May. Those four offerings are Enterprise AI Cloud, Enterprise Inference Engine, Inference as a Service, and Bare Metal AMD Instinct. Together, they are intended to provide what the companies call a complete, governed stack running from bare-metal compute through fully operated inference.

The companies framed this combination as an alternative to what they described as the bare metal model, in which enterprises assemble infrastructure from multiple vendors without a single operator accountable for end-to-end outcomes.

Rackspace and AMD indicated that the shift from AI experiments toward agentic workflows embedded in core enterprise systems is driving demand for the kind of governed, accountable infrastructure the collaboration is designed to deliver.

Executive Commentary and Commercial Model

Gajen Kandiah, chief executive officer of Rackspace Technology, said enterprises in regulated industries require AI infrastructure governed from the ground up, with one operator accountable for business outcomes rather than a collection of vendors each responsible for only a portion of the stack.

Kandiah said the collaboration delivers what he characterized as something the market has not previously offered: a governed AI stack with one accountable partner from silicon to outcomes.

Dan McNamara, senior vice president and general manager of Compute and Enterprise AI at AMD, said customers need infrastructure capable of delivering the right mix of accelerated and general-purpose compute for each workload as enterprise AI evolves.

McNamara said combining AMD's AI compute solutions with Rackspace's governed cloud operating model is intended to help regulated enterprises deploy high-performance AI infrastructure with openness, scalability, and accountability at enterprise scale.

Both companies said they will dedicate sales and marketing personnel to jointly identify and pursue customer opportunities across regulated industries. The agreement calls for each company to commit staff specifically to developing and engaging enterprise customers for AMD compute-powered infrastructure.

Market Positioning and Industry Context

Rackspace describes itself as the operator of a full enterprise AI stack spanning governed private cloud through AI inference and agents in production. The company's stated focus is on regulated and mission-critical industries where governance, data sovereignty, and uptime are, in its characterization, non-negotiable.

The Outcomes-as-a-Service model referenced in the company's description is built on secure infrastructure, data foundations, and what Rackspace calls forward-deployed engineering.

AMD describes its broader portfolio as including AI-optimized CPUs, GPUs, networking, and software. The company says its technology currently powers experiences across cloud and AI infrastructure, embedded systems, AI PCs, and gaming, and positions its full-stack AI solutions as designed for what it calls a new era of intelligent computing.

The two companies indicated that by jointly pursuing the regulated enterprise segment, they aim to establish a new category of managed enterprise AI infrastructure.

That positioning is tied directly to the governance requirements of industries such as healthcare, where compliance obligations and data sensitivity create barriers to adoption of standard public cloud or bare-metal configurations.

Power Your Data Center Pipeline With Intelligence That Spans Markets Worldwide

As demand for digital infrastructure accelerates across markets worldwide, the window for identifying and acting on high-value data center opportunities is narrowing. Are you confident your team has full visibility into every project that matters before your competitors do?

The Global Project Tracking (GPT) platform by Blackridge Research is purpose-built to give data center developers, contractors, and investors a structured, continuously updated view of the entire project lifecycle, from early planning signals through to completed facilities.

Whether you are monitoring hyperscale campuses, colocation builds, or edge deployments across regions, the platform consolidates critical intelligence into a single source of truth, helping your business prioritize pursuits and align resources with confidence.

• Upcoming Projects

• Tender Notices

• Contract Awards

• Projects Under Construction

• Completed Projects

See how the Global Project Tracking (GPT) platform by Blackridge Research can sharpen your data center strategy. Book a Free Demo with our team today.