Rebellions, a global provider of AI inference infrastructure, has announced a strategic collaboration with SK Telecom and Arm to develop next-generation AI infrastructure tailored for sovereign AI and telecommunications-focused data centers.
The partnership brings together expertise across the AI infrastructure value chain, with a focus on designing and deploying high-performance, energy-efficient systems capable of supporting growing demand in the AI inference market. Central to the initiative is the development of an AI server that integrates Arm’s upcoming data center-grade AGI CPU with Rebellions’ proprietary AI accelerators.
The server architecture will combine Arm’s AGI CPU—built on the Arm Neoverse CSS V3 platform—with Rebellions’ RebelCard accelerator. In addition to hardware integration, the companies will jointly develop the full software stack, including firmware, to ensure seamless interoperability and optimized performance.
Validation of the system will take place within SK Telecom’s operational AI data center environments. This phase will include testing for performance, scalability, and stability in real-world conditions, particularly for sovereign AI models and telecommunications workloads involving large-scale data processing. The collaboration may also explore running SK Telecom’s proprietary foundation model, A.X K1, on the newly developed infrastructure.
Industry observers note that sovereign AI—where nations or organizations maintain independent control over AI systems and data—is becoming a key priority amid increasing concerns around data security, compliance, and geopolitical considerations. This collaboration is positioned to address these requirements by delivering localized, customizable AI infrastructure solutions.
Following successful validation, the partners plan to expand into global markets, with a particular focus on Asia. Rebellions aims to strengthen its presence in the sovereign AI data center segment by offering tailored solutions to telecommunications operators and public sector entities requiring secure, independent AI capabilities.
Jinwook Oh, Chief Technology Officer at Rebellions, highlighted the role of the company’s RebelCard accelerator and full-stack software in supporting next-generation AI data centers. He described the collaboration as a “one-team” effort that could set a benchmark for AI-specialized infrastructure development.
Jaeshin Lee, Vice President and Head of AI Business Development at SK Telecom, emphasized the competitive advantage of combining inference-optimized infrastructure with the company’s proprietary AI model. Meanwhile, Eddie Ramirez, Vice President of Go-to-Market for the Cloud AI Business Unit at Arm, underscored the importance of CPUs in managing workloads across increasingly complex AI systems.
