Rackspace Technology and AMD have signed a definitive agreement for the phased deployment of 30 MW of AMD-based AI compute across Rackspace’s global data centers. The deal converts a May 7, 2026 memorandum of understanding into a binding contract and positions AMD as a strategic silicon partner in Rackspace’s governed AI stack.
Deployment is scheduled to begin in late 2026 and run through 2028.
- The agreement covers an initial 30 MW footprint dedicated to AMD compute, which Rackspace described as meaningful capacity for regulated enterprise workloads.
- The infrastructure will utilize AMD Instinct GPUs, including the MI355X and MI350P, along with future successor solutions, paired with AMD EPYC CPUs.
- The deployment will be integrated into Rackspace’s Enterprise AI Cloud architecture, which includes Enterprise Inference Engine, Inference as a Service, and Bare Metal AMD Instinct offerings.
- The partnership targets regulated industries, with particular emphasis on healthcare providers running clinical AI and large-scale inference workloads that require governance, data sovereignty, and high uptime.
- Both companies will dedicate sales and marketing resources to pursue enterprise customers in regulated verticals.
Gajen Kandiah, CEO of Rackspace Technology, said the collaboration delivers “a governed AI stack with one accountable partner from silicon to outcomes,” contrasting it with multi-vendor models common in the space.
Dan McNamara, Senior Vice President and General Manager of Compute and Enterprise AI at AMD, noted that the combination of AMD’s accelerated and general-purpose compute with Rackspace’s operating model is intended to help regulated enterprises deploy AI infrastructure with greater openness, scalability, and accountability.
Rackspace hosted an investor call today at 8:30 a.m. ET to discuss the agreement. No additional financial terms were disclosed.