Rackspace Technology, Inc.: AMD and Rackspace Technology Sign Definitive Agreement for Phased Deployment of 30 MW of AMD AI Compute
AMD and Rackspace Technology (RXT) signed a definitive agreement to phase in an initial 30 MW of AMD-based AI compute across Rackspace data centers, starting late 2026 through 2028, according to the companies. The full deployment would support regulated workloads. The stack includes AMD Instinct GPUs and EPYC CPUs, with sales and marketing resources dedicated to regulated industries. An investor call is scheduled for June 16.
New strategic compute partnership with quantified initial capacity (30 MW) and AMD as silicon-layer partner, supporting enterprise AI infrastructure demand narrative.
AMD signed a definitive agreement with Rackspace for phased deployment of an initial 30 MW AMD AI compute footprint across Rackspace data centers (late 2026–2028).
Moderately positive near-term sentiment for AMD, with follow-through dependent on deployment authorizations, pricing, and end-customer conversion.
Background
The release operationalizes a May 7, 2026 memorandum of understanding and frames the deal as a governed enterprise AI infrastructure category for regulated/sovereign environments.
Why it matters
A definitive agreement with a specified initial compute footprint (30 MW) and a multi-year deployment window (late 2026–2028) is a tangible catalyst, but the release repeatedly notes that deployment authorizations and commercial terms remain subject to further execution.
Market relevance
This is a concrete enterprise AI compute partnership announcement with quantified initial capacity and a near-term investor call catalyst, but with conditional commercialization details.
Market effects
Reinforces a trend toward governed/sovereign enterprise AI infrastructure and managed inference stacks, potentially supporting broader demand for AI compute supply chains.
No specific regional revenue split disclosed; rollout is described as across Rackspace’s global data centers.
Multi-year, global data-center deployment plan may strengthen enterprise AI infrastructure positioning for regulated workloads internationally.
Alternative perspectives
Because deployment authorizations, pricing, and financing are still subject to further agreements, the 30 MW figure may not translate into near-term revenue impact.
Traders may want to monitor whether the “initial 30 MW” converts into signed customer contracts and whether AMD GPU/CPU mix (Instinct GPUs and EPYC CPUs) drives margin and utilization assumptions.
Key entities
- companyAMD
AMD is the strategic technology partner at the silicon layer, supplying AMD Instinct GPUs and EPYC CPUs for the governed AI stack deployments.
- companyRackspace Technology
Rackspace is deploying the initial 30 MW AMD-based compute across its global data centers and positioning it within its governed AI stack.
- executiveGajen Kandiah
Rackspace CEO quoted emphasizing governed AI infrastructure with one accountable operator from silicon to outcomes.
- executiveDan McNamara
AMD SVP and GM Compute and Enterprise AI quoted on matching accelerated and general-purpose compute for workloads.
