Restore LLM Inference Capacity Quickly with Shadow Engine Recovery in NVIDIA Dynamo

NVIDIA Dynamo introduces shadow engine recovery, eliminating the need for cold restarts when LLM engine processes fail. This feature keeps a fully initialized shadow engine on the same GPUs as the active one, allowing for a quick takeover in seconds and minimizing service disruption. By sharing existing weights, it enables faster recovery while handling re-initialization in the background.