NVIDIA, Rebellions, and SK Telecom announced a collaboration to advance South Korea's sovereign AI infrastructure. Rebellions will connect its next-generation R200 neural processing units to NVIDIA's platform using NVLink Fusion. SK Telecom plans to deploy the combined systems in its AI data center.
Collaboration scope
The project has support from Korea's Ministry of Science and ICT. Rebellions will apply NVIDIA's MGX rack architecture to its system development. The companies plan to assign prefill work to NVIDIA GPUs and decode work to Rebellions NPUs.
Company background
Rebellions develops inference accelerators for enterprises and nations. The company has commercial deployments in government and enterprise environments. SK Telecom has operated as a telecommunications provider since 1984 and now builds AI capabilities across infrastructure, models, and services.
What the R200 brings
The R200 NPU is designed for low-latency and energy-efficient inference on large models. It will use NVLink Fusion for high-bandwidth, low-latency scale-up networking within a rack. The MGX rack design lets GPU and NPU systems share power, cooling, and space.
Rebellions expects the R200 systems to be ready for deployment in 2028. SK Telecom plans to verify the combined GPU and NPU setup at its AI data center. Korea's science ministry also plans to extend the cooperation to other Korean NPU companies.