Rackspace (RXT) Investor update summary

Strategic partnership and business realignment

• Signed a definitive agreement with AMD for a phased deployment of 30 MW of AI compute, leveraging AMD Instinct GPUs and EPYC CPUs, from late 2026 through 2028, targeting regulated industries such as healthcare, financial services, energy, and government.

• Unified company structure and integrated go-to-market strategy focus on governed enterprise AI, emphasizing accountability, operational expertise, and a model-agnostic orchestration platform.

• Partnerships with Uniphore, Palantir, Dell, VMware, and Rubrik complement the AMD agreement, creating a comprehensive, enterprise-grade AI stack and enhancing cyber resilience and compliance.

• Both companies will jointly dedicate sales, engineering, and marketing resources to pursue enterprise customers and accelerate AI adoption.

• Forward deployed engineering talent embedded with customers ensures outcomes and SLA accountability.

Financial and operational impacts

• Initial CapEx for the first deployment is estimated at $50 million–$100 million, with financing sources identified and no material impact expected this year.

• Workforce realignment plan includes up to 15% reduction, incurring $14 million–$19 million in one-time charges in 2026 and yielding $75 million–$85 million in annualized run rate savings, with savings reinvested into AI solutions and infrastructure.

• Deployment leverages existing global data center capacity, providing fixed cost absorption and capital efficiency, with no significant new construction required.

• Execution of deployments is subject to financing, operational, and legal conditions, with AMD holding a right of first refusal on capacity sales.

• Anticipated benefits and timelines are forward-looking and may be impacted by market, regulatory, or operational factors.

Product and service offerings

• Four integrated capabilities to be delivered: Enterprise AI Cloud, Enterprise Inference Engine, Inference as a Service, and Bare Metal AMD Instinct.

• The offering provides a governed stack from bare metal compute to fully operated inference, targeting enterprises moving from AI experiments to production workflows.

• Multi-tiered product evolution: from bare metal infrastructure to managed inference services and full-stack enterprise AI cloud.

• Ecosystem designed to accelerate enterprise AI adoption and deliver value across infrastructure, platforms, and applications.