Nvidia said it is working with South Korean AI chip startup Rebellions and SK Telecom, a South Korean telecom operator, on sovereign AI infrastructure using NVLink Fusion and the MGX rack architecture.

Rebellions will connect its Rebel R200 processors to Nvidia’s platform.

SK Telecom plans to deploy the systems and link them to its AI models and services.

The companies also plan to split inference tasks between Nvidia graphics processing units and Rebellions chips, with Nvidia handling prefill and Rebellions handling decode.

South Korea’s Ministry of Science and Information and Communication Technologies supports the collaboration, which Nvidia said would support AI services tailored to the country’s language and industries.

Nvidia described NVLink Fusion as technology for semi-custom AI infrastructure.

D-Matrix adopted the rack-scale interconnect for its systems, showing support for third-party accelerators.

Rebellions focuses on AI inference systems. SK Telecom also launched AI data center infrastructure, but the reviewed materials do not disclose the deployment scale for this collaboration.

The announcement said Rebellions’ R200 systems are expected to be ready for deployment in 2028.

🔗 Source: NVIDIA

Recent Nvidia developments