# NVIDIA GPUs for prefill, Korean NPUs for decode: Rebellions plans R200 inference systems using NVLink Fusion and MGX racks, with SK Telecom planning deployment. The systems are expected to be ready in 2028.

Canonical: https://ournews.ai/live/nvidia-gpus-prefill-korean-npus-decode?lang=en
Published: 2026-11-09T14:13:44.788844+00:00
Updated: 2026-11-09T14:13:44.788844+00:00

Source: NVIDIA — https://nvidianews.nvidia.com/news/nvidia-rebellions-and-sk-telecom-announce-collaboration-to-advance-sovereign-ai-infrastructure-with-nvidia-nvlink-fusion
Source date: 2026-11-08
