← Shelf

Series Ongoing Updated to ch. 13 · 2026-10-05

Why Won't the Model Run? — Understanding AI Infra in Comics

Letian hits Run and the screen only answers: Out of Memory. Follow Letian, Azhe, Xiaoya and Keke into the inner world of large models — the city of parameters, the quantization pipeline, the VRAM workbench, the KV Cache archive — and learn, in comics, what actually decides whether a model can run.

#AI Infra#LLM#VRAM#Explainer Comic