NVIDIA partners with Rebellions and SK Telecom to advance South Korea's sovereign AI infrastructure buildout

Nvidia has partnered with South Korean AI chipmaker Rebellions and SK Telecom to build large-scale AI inference infrastructure using Nvidia's NVLink Fusion interconnect and MGX rack architecture, with Rebellions integrating its next-generation R200 neural processing unit into the Nvidia platform and SK Telecom deploying the systems in its AI data centers alongside its own models and services. The three companies will also explore a disaggregated inference architecture in which Nvidia GPUs handle prefill and Rebellions NPUs handle decode, aiming to improve throughput and energy efficiency. The effort, backed by South Korea's Ministry of Science and ICT, is positioned as part of the country's sovereign AI buildout, and R200-based systems are expected to be ready for deployment in 2028. The deal marks a notable endorsement of Nvidia's rack-scale platform by a local chip challenger and deepens the vendor's ties to Korean telecom infrastructure. What to watch next is whether the prefill-decode split delivers measurable efficiency gains and whether the 2028 timeline holds as Korea advances its sovereign AI data center plans.