I can not agree more with you. The author just has no clue on memory management. He is imaging that the virtual memory model can solve all cache pains. The reality is much painful.