NVIDIA Launches Optimized Deployment Recipes for Korea's Leading Open Models

NVIDIA has announced the release of optimized deployment recipes for Korea’s leading open models, including A.X K2 from SK Telecom and K-EXAONE 2.0 from LG AI Research. These recipes are designed to distribute models across multiple GPUs, requiring careful coordination of request routing and communication to maintain performance. NVIDIA is developing an open-source framework called Dynamo, aimed at deploying models in multi-node environments at data center scale. This framework manages routing and communication to maximize inference performance, facilitating faster production of Korea’s open models by developers.