Distributed LLM Inference Across NVIDIA Blackwell and Apple Silicon Over 10GbE

I connected an NVIDIA DGX Spark to a Mac Studio with a direct 10-gigabit Ethernet cable and split a...